text
stringlengths 8
5.77M
|
|---|
Total knee arthroplasty following high tibial osteotomy for osteoarthritis.
The authors assessed the effect of proximal tibial osteotomy on the results of a subsequent total knee arthroplasty. A retrospective, clinical and radiographic analysis was carried out between a study group of 39 patients with 42 total knee arthroplasties following osteotomy and a control group of 39 patients with 41 primary arthroplasties. Outcome was assessed using the Hospital for Special Surgery (HSS) knee score, pain, function, range of motion, and radiographic evaluation. The follow-up period averaged 37 months (range, 24-50 months). The study group had 88% good or excellent results using the HSS score, compared to 90% in the control group. Function and pain improved equally in both groups. The control group had, on average, 14 degrees greater range of motion (115 degrees v. 101 degrees) after arthroplasty. The control group had, on average, posterior inclination of the tibial plateau of 7 degrees before operation and 3 degrees after operation; the corresponding values in the study group were 2 degrees and 1 degree. Using the HSS score and pain and function as parameters, previous osteotomy does not seem to affect the outcome of total knee arthroplasty. Conversely, range of motion following arthroplasty appears to be less in those with prior osteotomy. In addition, a high tibial osteotomy may alter the inclination of the tibial plateau.
|
The latest tale in the Torchwood monthly range is out now. Toshiko Sato was always considered the brains of the Torchwood team - but that might be a problem when you're dealing someone who can explode your mind†¦
Torchwood: Instant Karma is out now.
Imagine. All those people. The ones who make each day that little bit harder. That little bit more unbearable.
Imagine if you could silence them. Just by looking at them.
I mean, just imagine. If you could do that. To all the people who annoy you.
Would you do it?
Producer James Goss tells us more: “In Instant Karma, Toshiko discovers a group with superpowers. Because it's Torchwood they're operating out of a community centre, and they're not using their powers for good. Imagine if you could make the heads explode of people who annoy you. Would you use that power? Well, would you?!”
You can pick up Torchwood: Instant Karma now for £7.99 on download or £9.99 on CD – don’t forget that all CD purchases unlock a download option from the Big Finish website and on the Big Finish app.
Or you can save money with a Torchwood bundle. Purchase stories 19–24 together for just £55 on CD or £45 on download.
Coming up next month in the Torchwood range is Deadbeat Escape, and Murray Melvin returns as Bilis Manger in his own sinister tale.
"Welcome to the Traveller's Halt. You're in luck – we have one room left. My name is Bilis Manger. Let me show you to your room.”
Out for release tomorrow, Torchwood One: Machines sees Yvonne Hartman and Ianto Jones in Torchwood One. An old threat is returning to London, and this time it’s got wifi…
Torchwood One – we defend the British Empire. We keep everyone safe. For your own good.
Something's happened. Something's twisted its way into Torchwood One. Something old. Something which should have been forgotten, wiped, abandoned. Something which has been waiting. Something which knows how the world should be ruled, how people should be used.
When humanity's threatened, Yvonne Hartman is required.
|
Find local Stonemasons in St Columb, Cornwall
MyBuilder has thousands of local and reliable Stonemasons in St Columb. We screen our trade members and every job is up for review. Post your job now to get quick responses from local Stonemasons in St Columb.
I have been in the building trade for over 15 years. Im
reliable hard working with a good sense of humour and
patience.My business is based around stonework and
plastering which consists of drystone walling and blockwork
to skimming pointing and rendering. Have a good...
Latest Stonemasonry feedback review:
Absolutely brilliant . Punctual and did a fantastic job . Did stonework and plastering which was finished beautifully . Highly skilled and really pleasant to have around . Will use again .
All aspects of brickwork and stone-masonry competently
undertaken. From newbuild to restoration and conservation,
cob, lime repair, architectural and decorative masonry
designed and made by hand. Many years experience in both
brickwork and stonework, very professional and...
Latest Stonemasonry feedback review:
I hired Ed to build a drystone plinth for me and from the outset he was enthusiastic and efficient. His can-do attitude was so refreshing.
I received a detailed sketch of what he was to build, and I am delighted to say the finished...
Hi, I am an experienced builder specialising in stone work
and roofing but am equally capable with all types of
renovations and extensions. Having been in the trade for
over 15 years and running my own business I understand the
importance of customer service, reliability ...
Latest Stonemasonry feedback review:
Roland repaired my gable end as required. I have a lot of comments saying how nice it looks from my neighbours so very pleased.
Traditional Stone Masonry.
All aspects of masonry undertaken.
Fully qualified and time served mason.
Emphasis on quality of workmanship and reliability.
Latest Stonemasonry feedback review:
Gareth was really lovely to deal with - friendly, helpful & professional. They completed the work on time and worked without disturbing us. The main damage was to the window roof & this has been repaired well, my only slight complaint...
Established for over 35 years.
We are family run, a father and son team consisting of a
Stonemason and a Bricklayer. City and guilds qualified and
insured, clients include the National Trust and Kingswear
Parish Council.
Committed to produce an end product that is long...
I am a skilled stonemason with more than 25 years experience
in the building trade. I'm self-employed and specialise in
stone masonry, having spent 15 years in France working with
local artisans on a diverse range of projects. I have
recently returned to England and I'm...
Latest Stonemasonry feedback review:
Stone cladding of existing curved retaining wall and creation of additional wall. Mark has done an excellent job and we would strongly recommend him. Very friendly and reliable.
Very experienced in all aspects of stonework, hard
landscaping and bespoke works.Have something built
especially for yourself,your ideas or we can help you with
your design.
Latest Stonemasonry feedback review:
Excellent experience! Jason was quick to respond to my post on My Builder he communicated effectively through out the process and always did exactly what he said he would. He completed the job on time and to the agreed budget. I would have...
Hi I'm a fully qualified mason who undertakes New Builds -
Extensions - Specialising also in Stonework.
I have lots of previous jobs with photos and recomandations
from happy customers over the years who have said are happy
for any of my new customers to see for themselves...
Latest Stonemasonry feedback review:
Very pleased with the rebuilding of the stone wall. A very fair price. Pleasant and obliging workmen. Site left neat and tidy. Would recommend.
G p building services is a family run business with a
trusted established reputation in and around the local area,
with customer satisfaction our top priority. We offer a wide
variety of services to help you make your house a home. All
jobs in and around your house or...
Latest Stonemasonry feedback review:
Very pleased with the end result. Several comments of approval from passers by. Graham is hard working and very pleasant.
Hi My names ben potter ,Bpbuilders is a local building
company covering the southwest for 12 years with excellent
feedback. We are fully qualified and insured, we undertake
all building work and plastering and lime work we don't
leave the job unless the customer is 100 %...
Latest Stonemasonry feedback review:
Ben and Billy did an excellent job building a stone wall for us they arrived on time each day and completed the work to schedule despite the terrible weather. I would certainly recommend them in the future and we have already asked Ben to...
Hello potential customers , Stone&Wood is a small and local
family business comprised of two brothers, Ryan & Ashley
Mason. We are hard working, friendly, conscientious and
passionate about our work.
Ryan is a carpenter/joiner.
Ashley is a bricklayer/stone mason.
We are...
Latest Stonemasonry feedback review:
Ashley & Ryan are two really solid guys who will not disappoint you.
They have been really committed to producing a piece of work that is both solid and beautiful.
I am a builder with over thirty five years experience, I sub-contracted...
I have been a builder for 25 years working in the torbay
area but mostly in the south hams area building one off
multi million pound houses . My work range is anything from
extensions ,stone and brickwork,slabbing and landscaping
,driveways etc
Latest Stonemasonry client reference:
Sam re-built the stone around 2 doorways in our barn conversion. His work is excellent, I would have no hesitation in recommending him. He is punctual, polite & tidy in his work. He always turned up when he said he would & completed the...
I specialise in traditional methods of repair. These methods
and the materials I use are the most beneficial for repairs
to older buildings. Lime rendering, pointing and plastering
and the repair of lath walls and ceilings are my main work.
I also undertake cob and stone...
Latest Stonemasonry feedback review:
I am so happy with the work that has been done. My exterior has been transformed with lime mortar pointing and looks really smart. Internally the render has been hacked off and we were going to re-render with lime mortar but the granite...
|
Celebration Condominiums
05May
Celebrations condominiums at Bridlewood Mall
This slideshow requires JavaScript.
It where you want to live, shop, enjoy and celebrate life to its fullest. Set within GTA at Warden and Finch. With the convenience of living steps to Bridlewood Mall, the TTC at your door step, Celebration is the place to be…
|
The graphic design team has been incredibly busy over the past few weeks providing the finishing touches to many of our models
Posted by FW:ToW_Sushi on Oct 3rd, 2010
We have been extremely lucky of late to have an influx of such high quality modellers/texturists on the team. With such talent, we decided to make another look at our shiplist and revamp some of the models we thought needed some love. If you've been following this mod in the past few months, you've seen the incredible contributions from Kalo, Dr. Knickers and Lord X. This news piece showcases the brilliant pieces of art completed by Sizer.
The first ship up to be redone was our I7 Howlrunner. The model we originally had fit the schematic seen below completely
That being said, we felt as a development team that the model itself was a little bland. Similarly to what we did with the E-Wing, we let Sizer used some of his artistic creativity and make the model much more pleasing to the eye.
As you can see the little bit of detailing and altering to the model really put it in line with some of our more stunning ships.
After seeing this fabulous work of art, I asked Sizer to take a look at our T-Wing. It was in the same boat as the I7 Howlrunner; canon model but bland.
Given the same artistic freedom he used with the I7, Sizer created this:
The Corellians have Sizer to thank for the wonderful remodelling of their core fighters!
And, saving the best for last, Sizer also went to work on our Escort Carrier. I was so pleased with the results!
Using this boring schematic:
Sizer created this:
As you can see, the graphic design team is absolutely top notch and we all are lucky to have them devote their time and efforts on this mod.
The last thing that I wanted to show our loyal fans was regarding a new infocard system. For those of you unfamiliar with Freelancer, infocards contain statistics about the vessel as well as background information regarding the development and purpose of the ship. Currian, our lead infocard writer, has really gone above and beyond my expectations with the creation of some of the most informative and interesting infocards I've seen in Freelancer. As well, with Adoxa's hack, he's created a very user friendly system which divides the information into background and technical specifications. Below is a video demonstration of the new infocard system:
As always any comments, suggestions or questions are always welcomed!
----------------------------------------------------------
|
One of the wonderful aspects about the Marvel casting process is how the studio fills ancillary roles with seasoned actors who usher in a wealth of nostalgia. So, the studio thrilled audiences with Robert Redford uttering, “Hail Hydra” as Alexander Pierce in Captain America: The Winter Soldier. There’s also the sly tactic of casting James Spader’s voice in Avengers: Age of Ultron to give that film a Pretty in Pink edge or inserting Michael Douglas into Ant-Man. The studio’s been hunting for such an actor to play Star-Lord’s estranged dad in Guardians of the Galaxy 2, and for awhile, the rumors pointed towards Kurt Russell, but the actor gave a forthright answer that indicated no done deal.
Now another pretty face is allegedly punching his way into an uncertain Guardians 2 role. Would you believe Sylvester Stallone was spotted near the set? If this story from New York Daily News is to be believed, Stallone could be the latest Marvel recruit. The paper published photos, which allegedly show Stallone with a very important script in hand, and there are more bits of evidence:
Sylvester Stallone was caught on camera Friday heading to the set of Guardians of the Galaxy Vol. 2. He is pictured holding a script, likely evidence of a role in the sequel to the 2014 blockbuster. Stallone, 69, was spotted leaving an Atlanta hotel where the movie’s cast stays, a source told the Daily News. The Creed star then got into a car usually assigned to Chris Pratt, the star of the movie.
Naturally, Marvel is keeping the intrigue alive and hasn’t responded to the paper’s inquiry. Nor has anyone, including Stallone, actually said anything about popping into the movie. He could simply be staying at the same hotel and climbed into the wrong car. It could happen, and the paper indicates that Stallone could merely be pulling a cameo with Russell stepping into the coveted daddy role. (Would Stallone even need a full-on script for a cameo?) Any role would be a treat for both fans of the actor and Marvel. Stallone’s career has hit a new stride after winning a Golden Globe, and the offers are likely rolling in. Perhaps he wants to move past his typical Stallone-esque roles by putting a ring on it with Marvel.
(Via New York Daily News)
|
The Family Secret
The Family Secret is the title of two separate unrelated films:
The Family Secret (1924 film)
The Family Secret (1951 film)
|
Covered versus uncovered self-expandable nitinol stents in the palliative treatment of malignant distal biliary obstruction: results from a randomized, multicenter study.
Covered biliary metal stents have been developed to prevent tumor ingrowth. Previous comparative studies are limited and often include few patients. To compare differences in stent patency, patient survival, and complication rates between covered and uncovered nitinol stents in patients with malignant biliary obstruction. Randomized, multicenter trial conducted between January 2006 and October 2008. Ten sites serving a total catchment area of approximately 2.8 million inhabitants. A total of 400 patients with unresectable distal malignant biliary obstruction. ERCP with insertion of covered or uncovered metal stent. Follow-up conducted monthly for symptoms indicating stent obstruction. Time to stent failure, survival time, and complication rate. The patient survival times were 116 days (interquartile range 242 days) and 174 days (interquartile range 284 days) in the covered and uncovered stent groups, respectively (P = .320). The first quartile stent patency time was 154 days in the covered stent group and 199 days in the uncovered stent group (P = .326). There was no difference in the incidence of pancreatitis or cholecystitis between the 2 groups. Stent migration occurred in 6 patients (3%) in the covered group and in no patients in the uncovered group (P = .030). Randomization was not blinded. There were no significant differences in stent patency time, patient survival time, or complication rates between covered and uncovered nitinol metal stents in the palliative treatment of malignant distal biliary obstruction. However, covered stents migrated significantly more often compared with uncovered stents, and tumor ingrowth was more frequent in uncovered stents.
|
Q:
Accidentally installed two versions of git
I originally had git installed via Homebrew, but on this machine I forgot to setup the bash completion. So when I attempted to download the git source I accidentally installed a second version of git (1.7.3) which overrode 1.8.0.
Obviously I should've been more careful. So, I attempted to uninstall that version of git in hopes the version installed with Homebrew would just work, but it's not.
So question is, how do I uninstall all versions of git, and then re-install the latest version with Homebrew?
I tried
rm -rf /usr/local/bin/git
rm /etc/paths.d/git
But after the first line, I just got: No such file or directory
I also uninstalled git through brew and then reinstalled it, but when I do:
git --version
I get nothing. :/
Any suggestions?
A:
Have you checked your $PATH to see if it's correct set? Remember /usr/local/bin should come before /usr/bin when using Homebrew (or whatever path you have set for your installation of Homebrew.)
Example: (this should be in your .profile file, or .zprofile if you use Zsh)
PATH=/usr/local/bin:$PATH
export PATH
This way, the Git version installed via Homebrew should be detected first before any system installations.
But we can easily debug this, if you type the following command into your shell and post the output:
$ echo $PATH
Good luck!
|
Sarbat da bhala
Sarbat da bhala is the final term in the Sikh prayer called the Ardas. The term outlines one of the most important of Sikh principles and is a point that is repeated in the Sikh Scriptures. The Sikh concept of sarbat da bhala which means "blessings for everyone" or literally "may everyone prosper". This statement is repeated by all practising Sikhs at least twice daily as part of their Nitnem (daily prayers). This concept is central to Sikhism and forms a very important and essential role in the religious philosophy of the Sikh Gurus.
The Ardas ends with the line:
To put this in very concise and clear language, the term invokes the Sikh to request, pray and ask the Almighty for the well-being of all of humanity, prosperity for everyone in the worldwide community, global peace for the entire planet.
A Sikh selflessly prays daily for "all to prosper"'. This gesture comes from the clear and pure teaching of Gurbani (Sri Guru Granth Sahib) and forms the Gurmat code of conduct. Gurbani tells us that there are "no others", "there is only One". The same "One God resides within all". We are "all the children of that One God". As the potter makes pots of different forms and colors from the same basic clay; and as the goldsmith moulds jewellery of various types, colors and shapes from the same single homogeneous material, gold; similarly, we are all born of the same One Light: "There is no difference".
Accordingly, by instruction from the Gurus, the true Sikh begs for "the good for all beings" of the world. Because he knows that within every one of us resides the same One God – the Sikh should realise that he is "one with the rest".
Category:Sikh scripture
Category:Sikh prayer
|
The computing power needed to replicate the human brain’s relevant activities has been estimated by various authors, with answers ranging from 1012 to 1028 FLOPS.
Details
Notes
We have not investigated the brain’s performance in FLOPS in detail, nor substantially reviewed the literature since 2015. This page summarizes others’ estimates that we are aware of, as well as the implications of our investigation into brain performance in TEPS.
Estimates
Sandberg and Bostrom 2008: estimates and review
Sandberg and Bostrom project the processing required to emulate a human brain at different levels of detail. For the three levels that their workshop participants considered most plausible, their estimates are 1018, 1022, and 1025 FLOPS.
They also summarize other brain compute estimates, as shown below (we reproduce their Table 10). We have not reviewed these estimates, and some do not appear superficially credible to us.
Drexler 2018
Drexler looks at multiple comparisons between narrow AI tasks and neural tasks, and finds that they suggest the ‘basic functional capacity’ of the human brain is less than one petaFLOPS (1015).
Conversion from brain performance in TEPS
Among a small number of computers we compared , FLOPS and TEPS seem to vary proportionally, at a rate of around 1.7 GTEPS/TFLOP. We also estimate that the human brain performs around 0.18 – 6.4 * 1014 TEPS. Thus if the FLOPS:TEPS ratio in brains is similar to that in computers, a brain would perform around 0.9 – 33.7 * 1016 FLOPS. We have not investigated how similar this ratio is likely to be.
|
Soarin' sure looks good in uniformAlso, as someone below me said, it does sorta look like a TSA uniform!Good thing he isn't managing body search or item confiscation!
|
Do you believe what one person can achieve, others can achieve as well? How about if you knew that a whole group of people did something as simple as visualizations to achieve their goals that you could too?
Belief is a theme I teach in my Yoga for Golf classes and I use two examples to bring home the point that Practice in golf does not and should not only include physical practice.
The first example I use for my students comes a well-documented Soviet experiment using 4 matched groups of world class athletes.
The experiment was to put these groups prior to the 1980 Olympics. The experiment was for their training regimes. This is how it was devised:
Group 1 trained only physically,
Group 2 trained physically 75% of the time and mentally 25%
Group 3 divided physical and mental training equally
Group 4 put 25% into physical and 75% into mental.
The athletes were evaluated and Group 1 doing only physical training showed the least improvement in performance. The group that showed the most improvement was the group that did the most mental training. I think this shows guts and ingenuity to experiment with training regimes before the Olympics. The same as the results can start working for you immediately even if you are up against a pressure situation like a tournament, club championship, golfing with a big client.
The next example comes from an almost unbelievable story I found in one of the “Chicken Soup for the Soul Book” and this story is also retold slightly different in Denis Waitley’s book “The Psychology of Winning (no affiliate).
A man shot a 74 by playing 18 holes in his mind
Major James Nesmeth had a dream of improving his golf game- and he developed a unique method of achieving his goal. Until he devised this method, he was just your average weekend golfer, shooting in the mid – to low – nineties. Then, for seven years, he completely quit the game. Never touched a club. Never set foot on a fairway.
Ironically, it was during this seven-year break from the game that Major Nesmeth game up with his amazingly effective technique for improving his game- a technique we can all learn from. In fact, the first time he set foot on a golf course after his hiatus from the game, he shot an astonishing 74! He had cut 20 strokes off his average without having swung a golf club in seven years! Unbelievable. Not only that, but his physical condition had actually deteriorated during those seven years.
What was Major Nesmeth’s secret? Visualization.
You see, Major Nesmeth had spent those seven years as a prisoner of war in North Vietnam. During those seven years, he was imprisoned in a cage that was approximately four and one-half feet high and five feet long.
During almost the entire time he was imprisoned, he saw no one, talked to no one and experienced no physical activity. During the first few months he did virtually nothing but hope and pray for his release. Then he realized he had to find some way to occupy his mind or he would lose his sanity and probably his life. That’s when he learned to visualize.
In his mind he selected his favorite golf course and started playing golf. Every day, he played a full 18 holes at the imaginary country club of his dreams. He experienced everything to the last detail. He saw himself dressed in golfing clothes.He smelled the fragrance of the trees and the freshly trimmed grass. He experienced different weather conditions-windy spring days, overcast winter days, and sunny summer mornings. In his imagination, every detail of the tee, every individual greens.
He felt the grip of the club in his hands. He instructed himself as he practiced smoothing out this down-swing and the follow-through on his shot. Then he watched the ball arc down the exact center of the fairway, bounce a couple of times and roll to the exact spot he had selected, all in his mind. Visualization for golf is extremely powerful.
Not once did he ever miss a shot, never a hook or a slice, never a missed putt.
He took Twenty strokes off and Shot a 74.
This can help in you breaking 80 for the first time
Using visualizations is actually a very ancient practice used by yogis and other masters, when they would walk across hot coals and not burn their feet that is because they had first brought themselves to a relaxed state and saw themselves already accomplish this feat in their minds.
So although it is great to get out and Practice at the range evidence also reveals, when you Practice your best drives and putts in your mind it is helping you towards what’s known as the Reinforcement of Success. For the most powerful results it is best to recapture and recall your best golf in your mind just as you are going to sleep or before you get out of bed in the morning and by pre-playing every game you will be reinforcing future success. If you do it right in mental practice, you’ll do it right in life.
What is the best round you’ve ever played in your mind? Could it be breaking 80 for the first time? Visualize it before it happens!
Photo Credit: Dirk Hansen
|
It was a horrifying ordeal for the 71-year-old victim. She was carjacked and kidnapped and finally robbed Tuesday night. WPVI
MINQUADALE, Del. (WPVI) -- Police said two suspects are in custody in connection with the robbery and kidnapping of a woman in Delaware.The ordeal for the 71-year-old victim began at 5 p.m. Tuesday at a Wawa north of Wilmington and ended at another Wawa seven miles away.She was carjacked, kidnapped and robbed.But, it didn't take long for the police to grab a suspect.Police say 54-year-old Norris Farlow of New Castle is the man captured on surveillance inside the Wawa on Philadelphia Pike in Bellefonte.They say he pointed a gun at the victim in the store parking lot and got into her car with her."She was directed to various locations throughout the county, the city of Wilmington, then finally stopped at the Wawa at Memorial Drive, and that's when she handed over an undisclosed amount of cash to the suspect," said New Castle County Police Cpl. Tracey Duffy.Police say the suspect then left in another vehicle.Police sent out copies of the surveillance pictures of the suspect with patrol officers.Around 11:15 p.m. Tuesday, two officers stopped a car that rolled through a stop sign without stopping. Farlow was at the wheel."Through their observations and through the alerts they receive through our portal system, they were able to identify the driver of that vehicle as our suspect from the carjacking/kidnapping," said Cpl. Duffy.Inside the car, police say they found drug paraphernalia and the orange shirt Farlow was wearing in the surveillance video.Police said Farlow's accomplice, 51-year-old John Johnson, was arrested on Wednesday.New Castle County credited the patrol officers with getting Farlow off the street as quickly as they did."It's great police work by those two officers from D Squad. And again, it's great police work, great investigation all the way around," said Duffy.Police say Farlow has a long arrest record. Last December he was arrested for leading police on a high-speed chase that ended in a crash.Both men are being held on multiple felony counts.
|
The present invention relates to a technology effectively applicable to a data processor having a clock pulse generation circuit and particularly to a data processor which operates at low voltages and consumes relatively large power.
Patent document 1 describes the microcomputer's power-on reset technology in accordance with a power-on sequence. Patent document 2 describes the microcomputer's clock generation circuit having variable frequencies. There is considered a case of changing a frequency and an operating voltage. When the clock signal frequency and the operating voltage are increased, an attempt is made to prevent malfunction by avoiding an instance of increasing the frequency before increasing the operating voltage. Patent document 3 describes that the microprocessor and the peripheral circuit completely operate on the time-sharing basis in consideration for an increase in the power consumption when clocks are simultaneously supplied to the microprocessor and the peripheral circuit.
[Patent document 1]
Japanese Unexamined Patent Publication No. Hei 5 (1993)-333963[Patent document 2] Japanese Unexamined Patent Publication No. Hei 7 (1995)-287699[Patent document 3] Japanese Unexamined Patent Publication No. Hei 9 (1997)-231195
|
QUICK NOTE: The good folks at XFINITY sent me deep into the Fijian wilderness to bring you an exclusive look at “Survivor: Ghost Island.” While I was there I conducted interviews with “Survivor” host Jeff Probst and the entire cast. I also captured exclusive photos and other behind-the-scenes tidbits. So, be sure to follow me on Twitter (@gordonholmes) for up-to-the-minute updates.
Name: Desiree Afuye
Age: 21
Hometown: Newark, New Jersey
Current residence: Brooklyn, New York
Occupation: Student
Gordon Holmes: One of your hobbies is debating.
Desiree Afuye: (Laughs) Yes.
Holmes: Is that on a debate team or with random folks?
Afuye: No, I just debate! I will come up to you and debate. I will debate from every different angle, even if I don’t believe in the angle. I just like debating.
Holmes: That could get you in some trouble out here.
Afuye: I know when to debate and when not to debate. (Laughs) I know when to shut up. I’m a pretty good reader of people, so I’ll talk to the person as I see them.
Holmes: You also describe yourself as a YouTube binger. What’s your go-to genre?
Afuye: I love beauty videos. Like Gigi Gorgeous, the black hair beauty videos…even though I have no hair it still fascinates me.
Holmes: So make-up application…
Afuye: Make-up application, hair application…everything to do with beauty.
Holmes: Why do you think you were cast?
Afuye: I’m really fascinating.
Holmes: Obviously.
Afuye: (Laughs) I have a story that I bring along to this season. I was homeless at one point. I’ve been through a lot at a young age. I’m only 21.
Holmes: You’re 21, but you feel like an old soul.
Afuye: Yes, I get that all the time. And I think it’s because I’ve overcome certain things.
Holmes: How is this show even a challenge after what you’ve been through?
Afuye: Yes! After what I’ve been through, this is nothing.
Holmes: They should have to survive what you’ve been through.
Afuye: Exactly!
Holmes: Are you going to lie about any part of your backstory?
Afuye: I might lie, people might think I’m older than I am. I might say I have kids. I could be doing this for them. (Cries) This is for my baby. I don’t have any kids, I’m a kid.
Holmes: And how many “kids” do you have?
Afuye: I might have two…or one.
Holmes: You might want to pick a number and stick with it.
Afuye: One would be more believable.
Holmes: Boy or girl?
Afuye: A girl.
Holmes: And what’s her name?
Afuye: My baby girl Christina.
Holmes: Aww…and how old is she?
Afuye: She’s two years old.
Holmes: That’s a rough age because they get into everything and they’re moody.
Afuye: I know…but it was so hard for me to leave her.
Holmes: Who is she with?
Afuye: My boyfriend, he’s taking care of her.
Holmes: I’m picturing you at the finale with that million-dollar check, and some jury member being like, “This is going to do so much for you and Christina.” And you having to be like, “Yeah…about that…”
Afuye: (Laughs)
Holmes: You’re obviously OK lying. But have you set any kind of ethical boundaries for yourself?
Afuye: I know what game I’m playing. In the jungle you’ve got to do what you’ve got to do. And I’m here to do what I’ve got to do. If that includes lying? I’m sorry. I don’t lie outside of the game, but a million dollars is on the line.
Holmes: You mentioned a boyfriend…
Afuye: I don’t have a boyfriend.
Holmes: What?!
Afuye: (Laughs) Damn, I got you too!
Holmes: Lie to them, Desiree! I’m the only friend you have out here!
Afuye: (Laughs)
Holmes: I don’t know what to believe.
Afuye: (Laughs) I’m good.
Holmes: Alright, liar. Is flirting on the table?
Afuye: Yeah. I will flirt. I will use anything as leverage. If being a woman gets me ahead? I am the woman of all women. I am her.
Holmes: I was going to ask if your make-believe boyfriend would be upset if you were flirting. Maybe he’ll be make-believe upset.
Afuye: I thought we were going off of fiction!
Holmes: Are you a “Survivor” fan?
Afuye: Kind of, sort of. I’ve seen season 16, jumped to season 20, season 28, 32, 31, 34, all the thirties.
Holmes: They’re keeping you in the dark as far as twists and whatnot. Any guesses?
Afuye: I really think it’s Beauty vs. Brains vs. Brawn. Just by looking at the people.
Holmes: Which one are you?
Afuye: I think I’m beauty.
Holmes: I think I’m all of the above.
Afuye: (Laughs)
Holmes: Alright, I’ll give you a “Survivor” situation. You tell me which one you would rather deal with.
Afuye: OK.
Holmes: Align with a Hillary supporter or a Trump supporter?
Afuye: A Hillary supporter because Trump supporters are idiots.
Holmes: Find a steal a vote advantage or an eliminate a juror advantage?
Afuye: I’d rather steal a vote. I will woo any juror.
Holmes: It’s for Christina.
Afuye: It’s for baby Christina, c’mon!
Holmes: Baby clothes do not grow on trees.
Afuye: She deserves a good future.
Holmes: Endurance challenge or word puzzle?
Afuye: Word puzzle.
Holmes: Align with a racist or a sexist?
Afuye: A sexist…because a racist…I don’t understand how you can be racist. Why would the color of my skin determine who I am as a person? That’s dumb. A sexist deals with your ego. You feel so macho. I can deal with that. I can break you down. I can make you feel miniscule.
Holmes: Sexists get married.
Afuye: Yeah.
Holmes: They pledge the rest of their lives to a woman.
Afuye: Exactly.
Holmes: Whereas a racist doesn’t want anything to do with the other race.
Afuye: It’s interesting.
Holmes: Voted out first or voted out before the final Tribal?
Afuye: Voted out before final Tribal. At least I can prove to myself that I can come this far.
Holmes: Align with an adulterer or a tax cheat?
Afuye: An adulterer because a tax cheat is about the money. Some people play to be Sole Survivor, I’m about the money. If your goal is the same as mine, I don’t know if I want to align with you.
Holmes: When I was writing these I was thinking, these are some horrible people I’m making these folks align with.
Afuye: They are!
Holmes: But then you get in front of the final jury and say, “This dude is a racist tax cheat.”
Afuye: (Laughs) I wouldn’t need Christina.
Holmes: Lose the family challenge or win and have to decide who to bring along.
Afuye: Win and have to decide.
Holmes: Get caught stealing food or hunting for an idol?
Afuye: These are hard questions…caught stealing food. Because I can get caught, then everyone’s mad at me. And if I can’t woo them back, I’ll be alone and can go find the idol.
Holmes: Align with a super fan or someone who doesn’t watch the show.
Afuye: Doesn’t watch the show, cause then I’m running you and you’re my pawn.
Holmes: A dance challenge or a karaoke challenge?
Afuye: Karaoke, I like singing.
Holmes: What’s your jam?
Afuye: Oh…right now…has to be this UK artist, Skepta. It’s called “Skepta’s Interlude.”
Holmes: Is that made up? Like your boyfriend?
Afuye: It’s real. I want him to be my boyfriend.
Holmes: When people meet you, what is their first impression?
Afuye: I suffer from resting bitch face.
Holmes: Lemme see.
Afuye: (Sits normally)
Holmes: …I can see it.
Afuye: (Laughs) I have to consciously tell myself to keep a little smirk on my face. So, when people meet me and I have my resting bitch face, they think, “She doesn’t want to talk to me. She doesn’t like me.” To avoid that I’m going to keep a smirk on my face. I want people’s first impression to be, “She’s really nice. She’s my homegirl.” I’m going to be their homegirl.
NOTE: The cast members do not know each other’s names at this point. I showed Desiree a picture of each of the competitors and asked her to share her first impressions.
LIBBY Afuye: She smiles too much. She is the devil in a red dress. ANGELA Afuye: She may be trustworthy. But, I have to keep her on a leash. CHELSEA Afuye: She’s lying. Everything she says is a lie. Her name may be a lie. JENNA Afuye: She suffers from resting bitch face worse than me. She doesn’t know how to control her face and show that she’s nice.
Holmes: You two could start a support group.
Afuye: (Laughs) Yeah, we should. GONZALEZ Afuye: I want to work with her. I don’t know if I can trust her, but I think I can finagle her. STEPHANIE Afuye: She’s another one that smiles a whole heap of a lot. She’s very happy and bubbly. She wears bright colors and I think that’s all a plot to throw you off of her conniving ways. MORGAN Afuye: She’s very strong. She’s going to be my shield. LAUREL Afuye: I’m not sure about her. I’ll have to speak to her. KELLYN Afuye: I think I can trust her. BRADLEY Afuye: His eyes tell me a different story. He smiles, but his eyes tell me he’s trying to stab me in the back. BRENDAN Afuye: He reminds me of Mr. President. He’s kind of perfect. I’ll use him as a shield to keep the eyes off of me. CHRIS Afuye: His ego is crazy. I can already tell he loves himself more than anything else in this world and he thinks he’s a god. SEBASTIAN Afuye: I like him. I can’t trust him. WENDELL Afuye: I may be able to work with him. JACOB Afuye: I feel like he’s a super fan. Any move or anything I tell him, he’d already know. JAMES Afuye: He’s another one that’s like Mr. Perfect. I can’t read him, so that scares me. He’d have to go. MICHAEL Afuye: He might get into a romantic relationship on the show. But he has to go. DONATHAN Afuye: I don’t think he can really play. I think he’s going to play it safe. DOMENICK Afuye: This guy, it’s hard for him to smile. First one to go.
Don’t miss the season premiere of “Survivor: Ghost Island” – Wednesday, February 28, 2018 at 8 pm ET.
Any Questions? Drop me a line on Twitter: @gordonholmes
|
Bushiroad has unveiled a new trailer for the upcoming TV anime series for their male band Argonavis. The series is officially entitled Argonavis from BanG Dream! Animation. Check out the new trailer here!
The anime series is set for an April 10 premiere.
It was also announced that besides Argonavis, the anime series will also be featuring GYROAXIA. The latter was first introduced as one of the male bands featured in the Argonavis from BanG Dream! AASide game. GYROAXIA will also be performing “SCATTER”, one of the opening theme songs for the series. The other opening theme song, “Hoshi ga Hajimaru” (The Stars Begin), will be performed by Argonavis.
Hiroshi Nishikiori (Azumanga Daioh, A Certain Magical Index) is directing the series. Hikaru Miyoshi (Kanshikan Tsunemori Akane) is the character designer for both the game and anime. Writing the scripts for both the game and the anime will be Nobuhiro Mori.
Argonavis is composed of voice actors Shūta Morishima (keyboard), Seiji Maeda (bass), Daisuke Hyuga (guitar), Shōhei Hashimoto (drums), and Masahiro Ito (vocals). Their first single “Goal Line” was released on February 20. The band held their first live last May 17 at the Maihama Ampitheater in Maihama, Chiba.
BanG Dream! is a multimedia franchise formed by Bushiroad in January 2015. The franchise is composed of releases in the form of manga, music CDs, live concerts, card games, a smartphone game named BanG Dream! Girls’ Band Party! developed by Craft Egg, and a TV anime series. The anime series has ran for two seasons so far; the first aired from January to April 2017, while the second ran from January to March 2019. A third season has also been announced for the series, which is set to air on January 2020.
Despite coming from the same franchise, Bushiroad president Takaaki Kidani has stated that the male side of BanG Dream! has a different setting from the franchise’s female bands (Poppin’ Party, Roselia, Afterglow, Hello Happy World!, Pastel* Palettes, Glitter*Green, RAISE A SUILEN) and is an altogether
Source: Anime News Network
|
I know people will say it is early days and all that but, I really feel from here on out every game is a vital game. We need to grind out points which means we need Ipswich match like performances and no more Sheff Wednesday performances. Personally hopeing for at least a 22W 11D and 8L season or better.
I know people will say it is early days and all that but, I really feel from here on out every game is a vital game. We need to grind out points which means we need Ipswich match like performances and no more Sheff Wednesday performances. Personally hopeing for at least a 22W 11D and 8L season or better.
COYW!
That's just not accurate really. After the 1st International Break last season Reading were 15th and 2 Points off the Play-Offs and ended up finishing 3rd, 10 Points ahead of 7th, despite the fact they had some absolutely horrendous performances/results throughout the season.
I know people will say it is early days and all that but, I really feel from here on out every game is a vital game. We need to grind out points which means we need Ipswich match like performances and no more Sheff Wednesday performances. Personally hopeing for at least a 22W 11D and 8L season or better.
COYW!
That's just not accurate really. After the 1st International Break last season Reading were 15th and 2 Points off the Play-Offs and ended up finishing 3rd, 10 Points ahead of 7th, despite the fact they had some absolutely horrendous performances/results throughout the season.
I understand that and really admire teams that do that but, at what point do we say every game matters? 10 games in 20 games in? For me it needs to be now. We need to say that every game from now matters. That every point matters. Otherwise at some point the season just slips away and we look back saying oh, that is when we should have really focused and made the effort.
For me the time is now if we are going to make a push for promotion. Now is when we say this is it and we fight and grind every point we can out of every fixture.
I know people will say it is early days and all that but, I really feel from here on out every game is a vital game. We need to grind out points which means we need Ipswich match like performances and no more Sheff Wednesday performances. Personally hopeing for at least a 22W 11D and 8L season or better.
COYW!
That's just not accurate really. After the 1st International Break last season Reading were 15th and 2 Points off the Play-Offs and ended up finishing 3rd, 10 Points ahead of 7th, despite the fact they had some absolutely horrendous performances/results throughout the season.
I understand that and really admire teams that do that but, at what point do we say every game matters? 10 games in 20 games in? For me it needs to be now. We need to say that every game from now matters. That every point matters. Otherwise at some point the season just slips away and we look back saying oh, that is when we should have really focused and made the effort.
For me the time is now if we are going to make a push for promotion. Now is when we say this is it and we fight and grind every point we can out of every fixture.
Agree.
To say otherwise is a bit like not bothering to put any effort into your career on the basis that you might win the lottery. Yes you might go on an exceptional surge at some point but it's unlikely so we cannot rely on it. Last season we still only scraped into the play-offs with a bit of luck with Leeds' capitulation. We need to aim higher this year.
I doubt that Jokanovic and the players think or say that games do not matter that much, because it's still early. They wanted the 3 points against Ipswich as much as against Sheffield Wednesday and as much as they will want them against Cardiff, etc. Obviously there will be ups and downs, good days and bad days, days where mistakes will be made or the opponent will be better, but that doesn't mean that the game does not matter for them.
MJG asked for ideas for similar projections. We have a thread started 114 days ago that has reached 405 pages. How many pages will it get to by the close of the transfer window in three and a bit days time? At the average rate of posting I calculate it should get to 417. But intuition suggests somewhere in the mid to high 420s more likely. MJG (and others) views on this weighty matter welcomed.
So if my bonce has read this accurately, we need 1.9 points per match on average to secure 6th. So are we looking at targeting on average 2 wins out of every three matches to ensure this target is reached. That's all very well, but 6th place does not guarantee promotion, and neither does 3rd, 4th and 5th. Top two has to be the target, nothing less, otherwise as much as a victorious trip to Wembley in the play offs is pure theatre. That aspiration could be blown out of the water by a referees contentious decision, and we know all about them already. So for me top two is the order of the day, nothing less, otherwise I shall never live to be 100. As I reckon I aged 10 years during the last seasons play offs.
So if my bonce has read this accurately, we need 1.9 points per match on average to secure 6th. So are we looking at targeting on average 2 wins out of every three matches to ensure this target is reached. That's all very well, but 6th place does not guarantee promotion, and neither does 3rd, 4th and 5th. Top two has to be the target, nothing less, otherwise as much as a victorious trip to Wembley in the play offs is pure theatre. That aspiration could be blown out of the water by a referees contentious decision, and we know all about them already. So for me top two is the order of the day, nothing less, otherwise I shall never live to be 100. As I reckon I aged 10 years during the last seasons play offs.
Absolutely top 2 should be the target and I can add another part to the table to show that. Ive gone for 6th as its the minimum required. As I have said elsewhere the table will settle down by the 15th game and we will have a better picture of whats required..
At same stage last season the top 2 ended up being the top 2, 2 of next 4 made the playoffs and the other 2 were in top 10.7th place was on target for 74 (And it ended up at 75). Well within the variance range I use which is + or - 3pts.
Long way to go, but we need to be (and we will) there or there about at end of Oct
So if my bonce has read this accurately, we need 1.9 points per match on average to secure 6th. So are we looking at targeting on average 2 wins out of every three matches to ensure this target is reached. That's all very well, but 6th place does not guarantee promotion, and neither does 3rd, 4th and 5th. Top two has to be the target, nothing less, otherwise as much as a victorious trip to Wembley in the play offs is pure theatre. That aspiration could be blown out of the water by a referees contentious decision, and we know all about them already. So for me top two is the order of the day, nothing less, otherwise I shall never live to be 100. As I reckon I aged 10 years during the last seasons play offs.
Absolutely top 2 should be the target and I can add another part to the table to show that. Ive gone for 6th as its the minimum required. As I have said elsewhere the table will settle down by the 15th game and we will have a better picture of whats required..
At same stage last season the top 2 ended up being the top 2, 2 of next 4 made the playoffs and the other 2 were in top 10.7th place was on target for 74 (And it ended up at 75). Well within the variance range I use which is + or - 3pts.
Long way to go, but we need to be (and we will) there or there about at end of Oct
|
Analysis: Company report pace frustrates accounting experts
NEW YORK (Reuters) - Author and lawyer Michael Young jokes about the days when it took more time to get some companies’ financial statements than it did for Columbus to discover America.
Alas, those days are still here.
In an age of lightning-fast stock trades and instant communications, yearly and quarterly financial reports seem stuck on an industrial-era pace that some say was obsolete decades ago.
“Financial reporting technology allows all sorts of possibilities about which we could not even fantasize back in the Great Depression when our periodic system was put in place,” said Young, a partner at law firm Willkie Farr & Gallagher.
Young and other accounting experts have long called for a move to real-time, online reports in lieu of quarterly earnings statements that investors now use to decide whether to buy stocks.
Cheap computing power has made real-time reports more technically feasible, but advocates still struggle to draw wide support for the idea.
CISCO PERFECTS VIRTUAL CLOSE
“There isn’t anything imminent out there that would change the financial reporting model that we have today to make it more of a real-time model,” said Tom Omberg, head of Deloitte & Touche’s financial accounting and reporting services. “A lot of issues would have to get worked out before that could become a reality.”
Yet others say real-time reporting is worth revisiting.
“Almost any company with rational management tracks its performance daily and makes critical business decisions on the basis of data and trends that should be distilled and published for public consumption,” said Harvey Pitt, former chairman of the Securities and Exchange Commission.
Cisco Systems Inc (CSCO.O), for example, has perfected a “virtual close,” the ability to put together a complete set of financial statements in just a few hours. Free, open-source software has allowed even much smaller companies to construct dashboard-like digital displays that continuously track sales, orders and other key figures.
That kind of data could help make the stock market much more transparent, said Paul Miller, a University of Colorado accounting professor.
“If I’m an investor, why can’t I go on the Web and see what the Wal-Mart sales are today?” Miller asked.
Instead, investors have to wait to find out what happened months ago.
“It’s a constant state of incomplete information and the consequence of that incompleteness is discounted stock prices, which means a higher cost of capital,” he added.
The SEC has been shrinking the time between financial reports since 2002. The largest companies, which once had 45 days to file quarterly reports, now have just 40 days, while their deadline for annual reports has been trimmed to 60 from 90 days. Small companies can still take 90 days to file annual reports — still longer than Columbus’s 70-day voyage.
A big hurdle in speeding things up more is figuring out how to audit real-time figures or the systems that produce them.
“If you wanted to go to a more real-time report, for example, a voluntary disclosure that companies could make to their shareholders, I don’t see how it could feasibly be done with an auditor’s report,” said Scott Whisenant, an accounting professor at the Bauer College of Business in Houston.
RETAILERS DROP MONTHLY REPORTS
Some companies fear that more frequent reports would keep investors too focused on short-term performance. Retailers including Wal-Mart Stores Inc (WMT.N), Starbucks Corp (SBUX.O) and Dollar General Corp (DG.N), which once reported monthly sales, have stopped that practice.
Eventually, some companies may embrace real-time disclosures to build more trust in the capital markets, said W. David Stephenson, a homeland security expert and author of Data Dynamite, a book about data sharing.
“I do think that quarterly reporting is going to become an artifact at some point and that it’s going to start with some companies — maybe banks — that have a particular credibility problem,” he said.
One change that may pave the way to more real-time reporting is an SEC requirement that companies tag their financial data with computer codes called XBRL, or extensible business reporting language, he said.
Those tags will help streamline financial reporting and allow auditors and regulators to check it more quickly.
“It would obviously be a headache for investors to try to process financial information on a daily basis, but what a real-time system would allow investors to do is step back and say to themselves: ‘What period of analysis for this particular company makes sense?’” said Young, the lawyer at Willkie Farr.
More real-time reporting might also result in smaller gyrations in a company’s share price because investors could adjust their expectations each day, he said.
“What we could do is get beyond the one-size-fits-all torture of having everything reported during earnings season,” he added.
|
Q:
Async play sound in javascript?
how can I load asynchronously a sound in javascript ?
I want to play in loop severals sounds ans calculus in parallel.
This is the timeline :
10 seconds 7 seconds 10 seconds
|--------------------|------------------|--------------------|--- etc
| Sound1 playing |Sound2 playing | Sound1 playing |--- etc
| Do a calculus |Do a calculus | Do a calculus |--- etc
Sound1 and Sound2 last less than 5 seconds
Calculs last 1 second.
How can i do that in javascript ?
I must use Workers in HTML5 ?
Thanks.
A:
Playing sounds asynchronously in JS is in fact pretty simple. You just have to create new Audios and play() them right away instead of play()ing global Audio objects. Here's an example:
function playSoundAsync(url){
new Audio(url).play();
}
So in your case, your code would look like this (with a hack involving setIntervals and setTimeouts for syncing up the sounds):
// For simplicity I'll use MP3 files readily available on the Internet
var sound1url = 'https://audio.code.org/win3.mp3';
var sound2url = 'https://audio.code.org/failure3.mp3';
var calculusUrl = 'https://audio.code.org/failure1.mp3';
setInterval(function() {
new Audio(sound1url).play();
new Audio(calculusUrl).play();
}, 17000);
setTimeout(function() {
setInterval(function() {
new Audio(sound2url).play();
new Audio(calculusUrl).play();
}, 17000);
}, 10000);
<button onclick="new Audio(sound1url).play();">Play sound1 individually</button>
<button onclick="new Audio(sound2url).play();">Play sound2 individually</button>
<button onclick="new Audio(calculusUrl).play();">Play calculus individually</button>
In addition, you might notice this when you click the Play ${sound} individually buttons rapidly: A new playback of ${sound} starts without waiting for or interrupting the current ${sound} playback! This allows you to create an acoustic mess like this:
var cacophony = setInterval(function(){new Audio('https://audio.code.org/win3.mp3').play();}, 25);
<button onclick="clearInterval(cacophony);">Click here to stop this cacophony</button>
|
Ty Linton6-2 • 235 • LinebackerMatthews, N.C. • Charlotte Christian SchoolEnrolled at NC State for the 2014 spring semester ... Originally signed with North Carolina in 2010, and was recruited by Dave Doeren when he was at Wisconsin ... Elected to sign a pro baseball contract with Arizona instead, as the Diamondbacks drafted Linton in the 14th round and offered him a $1.25 million dollar signing bonus ... Coached by Jason Estep at Charlotte Christian School ... Started on both sides of the ball and helped lead Charlotte Christian to the NCSIAA 11-man Division I championship game as a senior ... Named to the NCISAA Division I all-state team three times ... Recorded 12 sacks on defense ... Ran for 542 yards and seven touchdowns on 118 carries on offense ... Also led Christian to the baseball NCISSA 3A state championship as a junior ...Coming out of high school, was ranked as the No. 25 player in North Carolina and the No. 63 linebacker in the country by Rivals.com and was ranked as the No. 27 outside linebacker in the country by ESPN.com.
|
1. Field of the Invention
This invention relates to a valve driving device for an internal combustion engine, and more particularly to a type thereof including a hydraulic lifter in each of the rocker arms.
2. Discussion of the Background
In an ordinary internal combustion engine, it has been required to provide a valve clearance of a predetermined length for eliminating adverse effects caused by differences between the thermal expansions of the cylinder head, cylinder block and the like and those of the valve driving mechanism. In the situation where the clearance is not maintained correctly, noises and/or output loss due to the leakage of crude gas tend to occur during the operation. For preventing such a drawback, a hydraulic lifter has been proposed which maintains the valve clearance at zero during the operation of the valve driving mechanism while eliminating the noise and the output loss and stabilizing the operation of the mechanism. The present invention relates to a valve driving device for an automobile engine in which the hydraulic lifters are utilized.
Japanese Patent Disclosure No. 16,113/1978 discloses such a valve driving device for an internal combustion engine including hydraulic lifters. In this device, each hydraulic lifter is encased in a rocker arm, one end thereof receiving a load from a cam shaft while the other end thereof driving a valve stem. Each hydraulic lifter comprises a plunger slidably encased in the rocker arm so that one end of the plunger operates the valve stem, while the other end thereof forms a pressurized chamber, a reservoir formed in the body of the rocker arm, and a check valve which permits flow of a pressurized oil only in a direction from the reservoir to the pressurized chamber. An oil passage is further provided through a rocker shaft swingably supporting the rocker arm, so that an operative oil such as engine oil supplied from an oil pump is sent through the oil passage into the reservoir of the hydraulic lifter.
With the above described valve driving device, however, a large amount of bubbles tend to be mixed in the operative oil supplied into the reservoir through the oil passage formed in the rocker shaft at the time of engine start or abruptly increasing or decreasing the rotating speed of the engine. The bubbles are collected in an upper part of the reservoir, and when the amount thereof becomes excessive, the bubbles intrude into the pressurized chamber, losing the rigidity of the hydraulic lifter and entailing creation of impacting sound between the plunger and its stopper or between the valve and the valve seat.
|
// Copyright 2016, Google Inc.
// All rights reserved.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
// File detached comment 1
// File detached comment 2
// File leading comment 1
syntax = "proto3";
// Ignored detached comment
// Ignored package leading comment
package grpc.testing;
message Request {
}
message Response {
}
// ServiceA detached comment 1
// ServiceA detached comment 2
// ServiceA leading comment 1
service ServiceA {
// MethodA1 leading comment 1
rpc MethodA1(Request) returns (Response); // MethodA1 trailing comment 1
// MethodA2 detached leading comment 1
// Method A2 leading comment 1
// Method A2 leading comment 2
rpc MethodA2(stream Request) returns (Response);
// MethodA2 trailing comment 1
}
// Ignored ServiceA trailing comment 1
// ServiceB leading comment 1
service ServiceB {
// ServiceB trailing comment 1
// MethodB1 leading comment 1
rpc MethodB1(Request) returns (Response);
// MethodB1 trailing comment 1
}
// Ignored ServiceB trailing comment 2
// Ignored file trailing comment
|
The Lord Lieutenant of Ireland (also known as the Viceroy or in the Middle Ages as the Lord Deputy) was the head of England's (pre-1707) or Britain's (post 1707) administration in Ireland.
The office was originally the central focus of English/British administration in Ireland under the Lordship of Ireland (1171-1541), the Kingdom of Ireland (1541-1800) and the United Kingdom of Great Britain and Ireland (1801-1922). As the name suggests, the holder was in effect the King's representative; the word viceroy comes from the french vice roi or deputy king. Though earlier Lords Deputy had been Irish noblemen, from the Middle Ages, with the very odd exception, only English or British noblemen were appointed to the office.
The official residence of the Lord Lieutenant was the Viceregal Apartments in Dublin Castle. However from the late eighteenth century, the Lord Lieutenant lived for much of the year in the Viceregal Lodge (now Áras an Uachtaráin, the Irish presidential palace), a more private residence located in the Phoenix Park in Dublin. In later years, Lords Lieutenant only lived in the Castle during the 'Social Season' (early January to St. Patrick's Day (March 17), during which time they held social events; balls, drawing rooms, etc. Other summer or alternative residences used by Lord Lieutenant or Lords Deputy included Abbeyville in Kinsealy (now the home of former taoiseach Charles Haughey) and a Chapelizod House, in which the Lord Lieutenant lived while Dublin Castle was being rebuilt following a fire but which he left due to the building being haunted. Lords Lieutenant and earlier Lords Deputy sometimes also owned property in Ireland, in which they lived rather than in state residences. The Geraldine Lords Deputy, Gearoid Mór Fitzgerald[?] and Gearoid Óg Fitzgerald[?] being native Irish both lived in, among other locations, their castle in Maynooth[?], Co. Kildare. The Earl of Essex owned Durhamstown Castle[?] near Navan[?] in County Meath, a short distance from the residence of the Lord Bishop of Meath at Ardbraccan.
The Lord Lieutenant's government was not in any real way responsible to the Irish Parliament, prior to parliament's abolition thanks to the Act of Union passed in 1800. Nevertheless, he did hold a formal State Opening of Parliament[?], delivering his speech outlining his government policy programme from the throne on the dias in the Irish House of Lords[?].
By the mid 19th century, the Lord Lieutenant's role changed substantially. Though still the official representative of the sovereign, the day to day role of governing fell to the Chief Secretary for Ireland[?], who was in effect the prime minister of the British administration in Ireland. Many nineteenth century Lords Lieutenant were not even nominally members of the British Cabinet, while the supposedly more junior Chief Secretary usually was.
The office of Lord Lieutenant, like the English and British government in Ireland was generally unpopular with Irish nationalists, though it was supported with varying degrees of enthusiasm by the Irish unionist community. Some Lords Lieutenants did earn a measure of popularity in a personal capacity among nationalists. From the early nineteenth century, calls were made frequently for the abolition of the office and its replacement by a Secretary of State for Ireland. Though on one occasion, a Bill was even introduced by one government to make this change, the office survived right down until the end of British rule in Ireland.
Irish nationalists throughout the nineteenth and early twentieth centuries campaigned for a form of Irish self-government. Daniel O'Connell sought Repeal[?] of the Act of Union, with the re-establishment of a Kingdom of Ireland, while later nationalists like Charles Stewart Parnell sought a more moderate form of home rule[?]within the United Kingdom of Great Britain and Ireland. Both made clear however, that the office of Lord Lieutenant could not survive in a restructured system of Irish government.
The last of the four Home Rule bills, the Government of Ireland Act, 1920, did provide for the continuation of the office. The Act divided Ireland into two states, Northern Ireland and Southern Ireland. Two institutions were meant to join the two states; a Council of Ireland[?] (which was hoped would evolve into a working all-Ireland parliament) and the Lord Lieutenant who would be the nominal chief executive of both regimes, appointing both prime ministers and dissolving both parliaments. In fact only Northern Ireland functioned as a state, with Southern Ireland being replaced by the Irish Free State. The powers meant to have been possessed by the Lord Lieutenant were delegated by amendment to a new Governor of Northern Ireland[?], while the role of representative of the Crown in the Free State went to a new southern Governor-General. The Lord Lieutenantship as a result was abolished.
Lord Fitzalan was the last holder, with the office abolished in December1922.
By tradition the coat of arms of each Lord Lieutenant was displayed somewhere in the Chapel Royal in Dublin Castle; some were incorporated into stained glass windows, some carved into seating, etc. Dubliners noted that the last available space was taken by the last Lord Lieutenant, Viscount Fitzalan. Fitzalan was the first Roman Catholic appointed as a representative of the Crown since the Glorious Revolution that brought William and Mary to power in 1688.
|
Star Wars Commander
A Force-Infused Clash Of Clans
Supercell’s Clash of Clans has inspired a number of “homages,” but few have managed to be quite as sticky as the original. The latest challenger has the Force on its side, though.
Star Wars Commander gives players control of either the Empire or the Rebel Alliance during the Galactic Civil War. Players will build a base, train units, and base invasions against the computer or other players.
The game takes players across locales from Episodes IV - VI with either the might of Imperial AT-STs, AT-ATs, and other machines for war or the heroes of the Alliance like Han Solo and Chewbacca on your side.
Star Wars Commander is available for iOS devices now as a free download, with in-app purchases.
Facepalm. I used to love playing Star Wars strategy games on PC years ago: Rebellion, Force Commander, Galactic Battlegrounds, Empire at War... But this looks like such watered down trash. Thanks mobile gaming for helping to bring weak sauce versions of the game types I used to play. I have no interest in touching this on iOS even though its free to play... actually especially because its free to play.
Wow.. First came MOBAs...
Now what are they going to call this new genre...? gee...
MOBAs and games like Clash of Clans aren't my type of games.. Don't really get how people like these kind of games. :|
I'm upset with these latest Star Wars games. There are plenty of fun mobile games, but recently none of them have been Star Wars games. I hope that Star Wars will get its own, original mobile game sometime soon. Before DICE's Battlefront hits the scene. If not that, before Visceral and BioWare have released their Star Wars games.
|
@inject SignInManager<ApplicationUser> SignInManager
@{
var hasExternalLogins = (await SignInManager.GetExternalAuthenticationSchemesAsync()).Any();
}
<ul class="nav nav-pills nav-stacked">
<li class="@ManageNavPages.IndexNavClass(ViewContext)"><a asp-page="./Index">Profile</a></li>
<li class="@ManageNavPages.ChangePasswordNavClass(ViewContext)"><a asp-page="./ChangePassword">Password</a></li>
@if (hasExternalLogins)
{
<li class="@ManageNavPages.ExternalLoginsNavClass(ViewContext)"><a asp-page="./ExternalLogins">External logins</a></li>
}
<li class="@ManageNavPages.TwoFactorAuthenticationNavClass(ViewContext)"><a asp-page="./TwoFactorAuthentication">Two-factor authentication</a></li>
</ul>
|
September 21, 2018
Latest Research Hints at Predicting Autism Risk for Pregnant Mothers
In a number of recent studies, researchers at Rensselaer Polytechnic Institute have made enormous strides toward ASD prediction and diagnosis
Researchers at Rensselaer Polytechnic Institute—led by Juergen Hahn, professor and head of biomedical engineering—are continuing to make remarkable progress with their research focused on autism spectrum disorder (ASD). A recent paper authored by Hahn and Jill James from the University of Arkansas for Medical Sciences (UAMS) in the journal Research in Autism Spectrum Disorders discusses their work on predicting with approximately 90 percent accuracy whether a pregnant mother has a 1.7 percent or a tenfold increased risk of having a child diagnosed with ASD.
Currently there is no test for pregnant mothers that can predict the probability of having a child that will be diagnosed with ASD. Recent estimates indicate that if a mother has previously had a child with ASD, the risk of having a second child with ASD is approximately 18.7 percent, whereas the risk of ASD in the general population is approximately 1.7 percent.
“However,” said Hahn, a member of the Rensselaer Center for Biotechnology and Interdisciplinary Studies, “it would be highly desirable if a prediction based upon physiological measurements could be made to determine which risk group a prospective mother falls into.”
Hahn’s work in developing a physiological test to predict autism risk is part larger emphasis on Alzheimer’s and neurodegenerative diseases at the Center for Biotechnology and Interdisciplinary Studies, and an example of how the interdisciplinary life science and engineering interface at Rensselaer offers new perspectives and solutions for improving human health.
In this study, metabolites of the folate-dependent transmethylation and transsulfuration biochemical pathways of pregnant mothers were measured to determine whether or not the risk of having a child with autism could be predicted by her metabolic profile. Pregnant mothers who have had a child with autism before were separated into two groups based on the diagnosis of their child whether the child had autism or not. Then these mothers were compared to a group of control mothers who have not had a child with autism before.
The researchers concluded that while it is not possible to determine during a pregnancy if a child will be diagnosed with ASD by age 3, they did find that differences in the plasma metabolites are indicative of the relative risk (18.7 percent vs 1.7 percent) for having a child with ASD.
“These are exciting results as they hint at differences in some metabolic processes that potentially play a role in increasing the risk of having a child with ASD,” said Hahn.
In addition to the lead authors, Juergen Hahn of Rensselaer and Jill James of UAMS, this work included collaborators from Rensselaer, the University of Arkansas for Medical Sciences, and the MIND Institute at UC Davis.
This new research follows an earlier study published in 2017, which developed an algorithm based on levels of metabolites found in a blood sample that can accurately predict whether a child is on the autism spectrum. A follow-up study this spring was also highly promising in assessing whether a child is on the autism spectrum. These results have the potential for earlier diagnosis for ASD, and efforts are underway to develop a commercially available test based upon these findings.
Important research like this exemplifies the vision of The New Polytechnic, an emerging paradigm for higher education which recognizes that global challenges and opportunities are so great they cannot be adequately addressed by even the most talented person working alone. Rensselaer serves as a crossroads for collaboration—working with partners across disciplines, sectors, and geographic regions—to address complex global challenges, using the most advanced tools and technologies, many of which are developed at Rensselaer. Research at Rensselaer addresses some of the world’s most pressing technological challenges—from energy security and sustainable development to biotechnology and human health.
About Rensselaer Polytechnic Institute
Rensselaer Polytechnic Institute, founded in 1824, is America’s first technological research university. For nearly 200 years, Rensselaer has been defining the scientific and technological advances of our world. Rensselaer faculty and alumni represent 86 members of the National Academy of Engineering, 18 members of the National Academy of Sciences, 26 members of the American Academy of Arts and Sciences, 8 members of the National Academy of Medicine, 8 members of the National Academy of Inventors, and 5 members of the National Inventors Hall of Fame, as well as 6 National Medal of Technology winners, 5 National Medal of Science winners, and a Nobel Prize winner in Physics. With 7,000 students and nearly 100,000 living alumni, Rensselaer is addressing the global challenges facing the 21st century—to change lives, to advance society, and to change the world. To learn more, go to www.rpi.edu.
|
{
"version": 1,
"tags": [
{
"name": "a11y-compare-image",
"description": "`a11y-compare-image`\nLayers images over each other with a slider interface to compare them\n\nAttributes:\n\n * `active-layer` {`number`} - \n\n * `__lower` {`string`} - \n\n * `__upper` {`string`} - \n\n * `opacity` {`boolean`} - mode for the slider: wipe\n\n * `position` {`number`} - \n\n * `__markers` {`any[]`} - \n\nProperties:\n\n * `activeLayer` {`number`} - \n\n * `__lower` {`string`} - \n\n * `__upper` {`string`} - \n\n * `opacity` {`boolean`} - mode for the slider: wipe\n\n * `position` {`number`} - \n\n * `__markers` {`any[]`} - ",
"attributes": [
{
"name": "active-layer",
"description": "`active-layer` {`number`} - \n\nProperty: activeLayer"
},
{
"name": "__lower",
"description": "`__lower` {`string`} - \n\nProperty: __lower"
},
{
"name": "__upper",
"description": "`__upper` {`string`} - \n\nProperty: __upper"
},
{
"name": "opacity",
"description": "`opacity` {`boolean`} - mode for the slider: wipe\n\nProperty: opacity\n\nDefault: false",
"valueSet": "v"
},
{
"name": "position",
"description": "`position` {`number`} - \n\nProperty: position\n\nDefault: 0"
},
{
"name": "__markers",
"description": "`__markers` {`any[]`} - \n\nProperty: __markers\n\nDefault: "
}
]
},
{
"name": "a11y-compare-image",
"description": "`a11y-compare-image`\nLayers images over each other with a slider interface to compare them\n\nAttributes:\n\n * `active-layer` {`number`} - \n\n * `__lower` {`string`} - \n\n * `__upper` {`string`} - \n\n * `opacity` {`boolean`} - mode for the slider: wipe\n\n * `position` {`number`} - \n\n * `__markers` {`any[]`} - \n\nProperties:\n\n * `activeLayer` {`number`} - \n\n * `__lower` {`string`} - \n\n * `__upper` {`string`} - \n\n * `opacity` {`boolean`} - mode for the slider: wipe\n\n * `position` {`number`} - \n\n * `__markers` {`any[]`} - ",
"attributes": [
{
"name": "active-layer",
"description": "`active-layer` {`number`} - \n\nProperty: activeLayer"
},
{
"name": "__lower",
"description": "`__lower` {`string`} - \n\nProperty: __lower"
},
{
"name": "__upper",
"description": "`__upper` {`string`} - \n\nProperty: __upper"
},
{
"name": "opacity",
"description": "`opacity` {`boolean`} - mode for the slider: wipe\n\nProperty: opacity\n\nDefault: false",
"valueSet": "v"
},
{
"name": "position",
"description": "`position` {`number`} - \n\nProperty: position\n\nDefault: 0"
},
{
"name": "__markers",
"description": "`__markers` {`any[]`} - \n\nProperty: __markers\n\nDefault: "
}
]
}
],
"globalAttributes": [],
"valueSets": []
}
|
The five states of Central Asia — Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan, and Uzbekistan — mark 27 years of independence in 2018. In an interview with The Diplomat, Paul Stronski, a senior fellow in the Russia and Eurasia Program at the Carnegie Endowment for International Peace, fields the full gamut of questions about Central Asia, from the hottest topics — Uzbekistan’s reforms and Kazakh succession — to what the world is missing in its view of the region.
Arguably, the hottest topic in Central Asia of late has been Uzbekistan. What do you think the most important “reforms” have been? And alternatively, in what areas has Uzbekistan struggled (or not quite attempted) to make progress?
Uzbekistan’s most visible changes are to its foreign policy. Under the late President Islam Karimov, Uzbekistan had a policy of self-isolation. This was due to the country’s poor human rights record, its periodic desire to put distance between itself and its former colonial power Russia, and finally its reticence to work cooperatively with its neighbors. Karimov was a prickly leader and had a difficult relationship with all the other leaders of Central Asia, which impeded the ability of Central Asian states to cooperate.
President Shavkat Mirziyoyev’s willingness to work with his Central Asian counterparts has created tremendous opportunities for Central Asia to cooperate on a myriad of long-standing problems (water, energy, transportation, trade, and regional economic development). Borders are open, which is facilitating direct country-to-country Central Asian trade. Internationally, today’s Uzbekistan is courting the West; relations with Turkey are also improving. Russian President Vladimir Putin recently visited with a 1,000-strong delegation. Everyone is now interested in the new Uzbekistan. That is a big change.
The second area where there is promise is in economic modernization. Today Uzbekistan has declared itself open to the outside world. Mirziyoyev’s government has liberalized the visa regime and is courting Uzbek diaspora to return to help jumpstart the country’s socioeconomic development. Some are returning — a remarkable decision since many of these Uzbeks left the country for political or economic reasons years ago.
Enjoying this article? Click here to subscribe for full access. Just $5 a month.
The government clearly wants to improve the investor and business climate, and attract international businesses. Economic modernization is needed because the previous economic system was not sustainable, particularly with the Uzbek youth population growing so quickly. Uzbek youths have limited job prospects at home. This has forced many Uzbeks to move abroad as economic migrants. Economic modernization, however, will be hard. Even with diaspora returnees, Uzbekistan’s biggest problem now is that it lacks the human capital needed to implement these reforms.
We need to remember, however, that reform in Uzbekistan is not full-democratic reform. I have no expectation that the country will become a democracy anytime soon. The justice sector and security services were early targets of Mirziyoyev’s reforms, but those efforts were likely geared toward shoring up his political power and sidelining competitors as Mirziyoyev consolidated power during his first year in power. Other reforms have been slower and focused on narrow and localized issues. Uzbeks are now able to speak out about corruption in daily life — that’s a big and positive change. If Uzbekistan truly addresses this sort of corruption, it will lead to improvements in people’s everyday life. What is also interesting about domestic political reform is a greater openness to discuss formerly taboo topics and calls for localized officials to be more accountable to solving the people’s problems. If it leads to improvements in socioeconomic conditions, it an important step in improving human security. If economic modernization and greater accountability lead to less corruption, better economic opportunities, and a higher quality of life, it is positive and should be encouraged.
As someone who lived in Uzbekistan for several years, I am hopeful for the country. Yet, it is too early to give these reforms a full assessment. Old habits die hard and Mirziyoyev’s biggest challenge is breaking apart the old system without undermining his own legitimacy as someone who was a key participant in Karimov’s regime. Mirziyoyev’s government also has been lucky thus far. It has not faced any external or internal threats that would force it to react quickly. Its response to a crisis (a large protest at home or instability in a neighboring country) will likely be a significant test for whether Tashkent has turn the tide. We have not yet seen that sort of critical juncture.
Uzbek President Shavkat Mirziyoyev, in his first two years in office, made regional cooperation a centerpiece of his foreign policy. How has Kazakhstan, in particular, responded to this spark of initiative and drive for regional leadership in its poorer but much more populous neighbor?
So far, there is little competition between today’s Uzbekistan and Kazakhstan, although I certainly see it as a possibility. During the Soviet and pre-Soviet eras, Tashkent was the most influential city in Central Asia and Uzbekistan was seen as the regional leader. Uzbeks still see themselves as the regional leader, and this perception contributed to the difficult relationship between the Uzbek and Kazakh governments under President Karimov. The past two-and-a-half decades were lost decades for Uzbekistan, as the Kazakh economy grew and Astana became both the regional leader and an aspiring global power.
Kazakhstan now has the know-how and economy to lead the region; it is in a position to help Uzbekistan. But, Uzbekistan still has the strongest security forces and the largest population in the region. I think there is there is a lot of rationale for the two government to work together in the future rather than to compete. Together, they are a huge market that can attract outside investors; together, they have greater leverage in their engagement with outside powers than they do each on their own.
In fact, the two countries’ leaders seem to be on on similar schedules in courting the West. Nazarbayev came to Washington in January of 2018 to engage President Donald Trump directly, followed just a few months later by Mirziyoyev. The Kazakh leader made a similar visit to Brussels in October 2018 with his Uzbek counterpart reportedly planning a similar visit in early 2019. Given the growing lack of Western (especially American) attention to Central Asia, these dual visits appear to be well-timed to complement each other and to signal to the West not to disengage from the region. I do not know whether Astana and Tashkent are coordinating these visits, but the close proximity of these head of state visits to Washington and Brussels is certainly notable.
Diplomat Brief Weekly Newsletter N Get first-read access to major articles yet to be released, as well as links to thought-provoking commentaries and in-depth articles from our Asia-Pacific correspondents. Subscribe Newsletter
At 78, Kazakhstan’s President Nursultan Nazarbayev is Central Asia’s oldest leader. Among regional watchers, speculation about succession is a favorite parlor game so I’ll ask you to place your bet: Will Nazarbayev stand for election again in 2020 or do you think it more likely he’ll bow out with a nod toward a chosen successor?
I avoid answering questions like this. The only person who knows what Nazarbayev will do is Nazarbayev himself. I’m also not a betting person, so I won’t speculate on who might succeed him whenever he decides to step down. We all know the names of the potential successors, but there could be wild cards too. Nobody predicted Vladimir Putin would succeed Boris Yeltsin a year before that succession took place. Few thought dentist Gurganbuly Berdymukhamedov would become president of Turkmenistan before it happened. So, I am reticent to speculate on a succession scenario in Kazakhstan.
What I do know is Kazakhstan has a tradition of calling surprise snap elections and there is buzz right now about early elections. I would not be surprised if the presidential election took place in 2019 instead of 2020. Nazarbayev recently reshuffled his government and laid out a plan to improve socioeconomic conditions, and he has been traveling extensively across the country since the summer. Early elections look likely.
Enjoying this article? Click here to subscribe for full access. Just $5 a month.
Kazakhstan was where, in 2013, Chinese President Xi Jinping announced the land-based portion of what we now call the Belt and Road Initiative (BRI). As Chinese involvement in Central Asia has increased, has there been a rise in anti-Chinese sentiment in Kazakhstan or elsewhere in the region? If so, in what ways has that sentiment been expressed and do regional governments consider it a problem that needs addressing?
When you talk to Central Asian government officials, China is a perennial topic of discussion. Those officials are adamant that everything is fine with China. It may be, but sometimes they are too adamant in making that case. The tremendous influx of money from China certainly poses risks for Central Asia and China itself: growing debt burdens, the lack of transparency in many deals, economic degradation, and influx of Chinese workers in some places. Central Asian countries generally also are not looking beyond the BRI to develop more sustainable economies, which is a problem. BRI infrastructure will be great for Central Asia to become a transit region, but infrastructure construction does not create long-term jobs for Central Asians — a problem given expected population growth over the next decade.
As the Kazakh land protests of 2016 showed, China’s growing economic influence risks inciting popular backlashes. The USSR cultivated anti-Chinese sentiment during the Cold War and some of that Sinophobia lingers. China has been trying to expand its soft power in the region to break through those stereotypes. It provide scholarships and fellowships to promising Central Asian students or academics, offers all-expense paid tours of China to opinion-makers in the region, and fosters the study of Chinese in Central Asia.
However, China’s policies in Xinjiang risk undermining Chinese soft power in the region. Central Asians follow what is going on across the border. I was just in Kazakhstan and many Kazakhs expressed concern about the plight of the Uyghurs, and want their government to be more vocal, particularly since ethnic Kazakhs in China are impacted as well. But the government in Astana is wary of discussing the issue. It puts the Kazakh government between a rock and a hard place.
In economic terms, Kazakhstan is far and away the most economically prosperous Central Asian state, followed distantly by Uzbekistan and Turkmenistan and even more distantly by Kyrgyzstan and Tajikistan. While resource wealth explains much of this disparity between Central Asian states, how much disparity is there within these states between the rich and the poor? Do you think this economic inequality could have political consequences for the states of the region?
Kazakhstan now classifies as a middle income country, but many Kazakh citizens don’t feel the benefits of this economy. Wealth is not distributed evenly and corruption has a corrosive impact on the daily lives of people. This is a problem. Nazarbayev’s October 5 address to the nation lays out plans to address socioeconomic issues. It is a recognition of the fact that the wealth of Kazakhstan is not being felt by many of its citizens. There also is a huge gap between Kazakhstan’s boom towns (basically Astana and Almaty) and the rest of the country.
You see similar trends in Uzbekistan, Kyrgyzstan, and Tajikistan: big gaps between capital cities and secondary cities, an urban-rural divide, and collapsed social-welfare networks. This is forcing many Uzbeks, Kyrgyz, and Tajiks to migrate elsewhere (Russia, Kazakhstan, Turkey). There are even reports of Kazakhs now becoming migrant laborers in South Korea, although the high numbers of Kazakh migrants in the country may be exaggerated.
Socioeconomic disenfranchisement has led to political instability in many parts of the world, and it is certainly possible this could happen in Central Asia where the gaps between elites and the economically disenfranchised is probably higher than elsewhere. Given looming succession challenges in Kazakhstan and Tajikistan, this wealth gap could prove challenging to new leaders, who may lack the political legitimacy their predecessors enjoyed. I also am particularly concerned about Turkmenistan right now. The social contract between the government and people has collapsed and the Turkmen people are hurting. The country looks fragile from the outside.
Many international conversations about Central Asia focus almost exclusively on extremism or geopolitics with regard to Russia and China — as important as those topics may be, what do you think is missing from outside conversations about Central Asia?
Yes, there is too much focus on geopolitics in international conversations about Central Asia. Central Asian states like it that way because it is far easier for them to focus on this topic as opposed to more urgent regional social or economic problems. The focus on geopolitics also reflects a desire of the region to be relevant to outside players, particularly the West, which is not all that interested or invested in Central Asia despite its rhetoric.
What is missing from the discussion is how the region itself is changing demographically. Migration is a big issue in Central Asia. Everyone focuses on migrant laborers to Russia, but we are not discussing enough about why Central Asians are going there and what happens to them when they are there. Some are being radicalized while living abroad. This suggests that Central Asia’s high unemployment and inflation rates are not just economic issues, but security ones too.
Furthermore, the region’s demographic issues are not just about labor migrants to Russia, but to other countries as well (Kazakhstan and Turkey). This is even occurring across all sectors of society. The wealthier and better educated are also looking further west. In addition, urban centers are pulling Central Asian rural youth out of the countryside. This is depopulating many villages of the young generation. The rural poor come to the cities in search of economic opportunities they lack where they grew up. They often live on the outskirts of urban centers in poorly regulated and built suburbs, creating a new urban underclass.
Finally, I also think there is a skewed picture of religion in the region. We focus too much on the extremist threat, which generally gets overplayed by the media and local governments. But, we are not looking at how religion is changing in the region. The Central Asia region is becoming less secular than when I first started going there 20 years ago. This change is natural and not a problem, but the political elites generally remain highly secular, while there is greater interest in Islam among the general public. Rural to urban shifts in populations likely are impacting how religion is being practiced and understood in the cities too. It might be worth thinking more deeply about this issue, particularly as power will move to the post-Soviet generation in the coming decade. Some academics are looking at religion, but we need to think more about how this shift may change Central Asian government behavior and our policy options.
|
Q:
Only one record returned after union all
Using an SQL interface into an AS/400 (iSeries) DB2 database but running into and issue where pallet information could be stored in two tables, one for non-allocated stock (warpall) and one for allocated stock (warpalq).
I need to return the records from both tables where applicable.
warpalq only has a handful of fields in relation to warpall
One of the fields I need from warpall is also used to join another table, there is no relative field in warpalq.
I'm using a union to return the data from both tables in one result set, then joining the other tables.
The issue is, when joining the third table (warcmdt) on the result set, it has null values / 0 in the commpm and fdsccd fields as it has no data for that in the respective table.
Is there a way that I can pull in commpm & fdsccd back into the results set for row 2?
The crosslink would be pal#pm = pal#pq so there is a relative link.
Running the SQL below returns this result set.
select
clntwf, commpm, fdsccd, clsqwf, pal#wf, dtcdpm, clsswf, srb#pm, descsc,
cor#or
from
(select substr(warpall.clntpm,1,2) as clntwf, pal#pm as pal#wf, commpm,
substr(warpall.clsspm,1,2) as clsswf, warpall.clsqpm * 1 as clsqwf, srb#pm,
dtcdpm, srd#pm
from warpall
where
locnpm <> 'ASSEMBLED PALLET'
and commpm <> 'ASSEMBLED PALLET'
and clsqpm <> 0
and clntPM <> 'D2'
and pal#pm = '1005609592'
union all
select substr(clntpq,1,2) as clntwf, pal#Pq as pal#wf, '0',
substr(clsspq,1,2) as clsswf, clsqpq * 1 as clsqwf, '0', '0', '0'
from warpalq
where
clsqpq <> 0
and clntPQ <> 'D2'
and pal#pq = '1005609592'
) as t9
left outer join warcmdt as t3
on
t3.clntcd = t9.clntwf and t3.commcd = t9.commpm
left join
warclss as t4
on
t4.clsssc = t9.clsswf
left join
warohrh as t5
on
t5.clntor = t9.clntwf
and t5.srn#or = t9.srd#pm
order by pal#wf asc
A:
You could try using a Common Table Expression (CTE) like this:
with
pal (clntwf, pal#wf, commpm, clsswf, clsqwf,
srb#pm, dtcdpm, srd#pm) as (
select
substr(warpall.clntpm,1,2) as clntwf,
warpall.pal#pm as pal#wf,
warpall.commpm,
substr(warpall.clsspm,1,2) as clsswf,
warpall.clsqpm * 1 as clsqwf,
warpall.srb#pm,
warpall.dtcdpm,
warpall.srd#pm
from warpall
where
warpall.locnpm <> 'ASSEMBLED PALLET'
and warpall.commpm <> 'ASSEMBLED PALLET'
and warpall.clsqpm <> 0
and warpall.clntPM <> 'D2'
and warpall.pal#pm = '1005609592'),
alc (clntwf, pal#wf, commpm, clsswf, clsqwf,
srb#pm, dtcdpm, srd#pm) as (
select
substr(warpalq.clntpq,1,2) as clntwf,
warpalq.pal#Pq as pal#wf,
pal.commpm,
substr(warpalq.clsspq,1,2) as clsswf,
warpalq.clsqpq * 1 as clsqwf,
'0',
'0',
pal.srd#pm
from warpalq
join pal
on waarpalq.pal#pq = pal.pal#wf
where
warpalq.clsqpq <> 0
and warpalq.clntPQ <> 'D2'
and warpalq.pal#pq = '1005609592')
select
clntwf, commpm, fdsccd, clsqwf, pal#wf, dtcdpm, clsswf, srb#pm, descsc, cor#or
from
(select *
from pal
union all
select *
from alc
) as t9
left outer join warcmdt as t3
on t3.clntcd = t9.clntwf
and t3.commcd = t9.commpm
left join warclss as t4
on t4.clsssc = t9.clsswf
left join warohrh as t5
on t5.clntor = t9.clntwf
and t5.srn#or = t9.srd#pm
order by pal#wf asc
This lets you reuse the pal CTE in multiple places without redefining it. That means you can join it in to the alc CTE to get commpm which is used later to join t9 with warcmdt. Now you shouldn't get that null. I also brought srd#pm into the alc CTE from warpall because that is used in the join between t9 and warohrh. If this is incorrect, you can replace pal.srd#pm with '0' in the alc CTE.
|
/*******************************************************************************
* Copyright (c) 2019, 2020 Eurotech and/or its affiliates and others
*
* All rights reserved. This program and the accompanying materials
* are made available under the terms of the Eclipse Public License v1.0
* which accompanies this distribution, and is available at
* http://www.eclipse.org/legal/epl-v10.html
*
* Contributors:
* Eurotech - initial API and implementation
*******************************************************************************/
package org.eclipse.kapua.service.device.management.exception;
import org.eclipse.kapua.KapuaErrorCode;
import org.eclipse.kapua.service.device.registry.Device;
import org.eclipse.kapua.service.device.registry.connection.DeviceConnectionStatus;
/**
* @since 1.1.0
*/
public enum DeviceManagementErrorCodes implements KapuaErrorCode {
/**
* The {@link Device} is not {@link DeviceConnectionStatus#CONNECTED}
*/
DEVICE_NOT_CONNECTED,
/**
* An error occurred when sending the {@link org.eclipse.kapua.service.device.management.message.request.KapuaRequestMessage}.
*
* @since 1.1.0
*/
SEND_ERROR,
/**
* A response as not been received within the given timeout.
*
* @since 1.1.0
*/
TIMEOUT
}
|
Q:
Pooling issue: Item borrowed more than once
I have a utility method (= a static one) that I call a lot, which uses a java.util.regex.Matcher. As the regex passed is reused a lot, I try not to compile it every time, so I keep it in a Map where the key is the regex, and the value is a List of Matcher objects (so that each thread gets it own Matcher instance).
How is it that the following code snippet manages to return the same Matcher twice... sometimes?
import java.util.HashMap;
import java.util.HashSet;
import java.util.Map;
import java.util.Queue;
import java.util.Set;
import java.util.concurrent.ConcurrentLinkedQueue;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
public class MyTest {
private static final Map<String, Queue<Matcher>> matchers = new HashMap<String, Queue<Matcher>>();
private static Set<Integer> duplicateHunter = new HashSet<Integer>();
private static Matcher getMatcher(String regexp, String value) {
Queue<Matcher> matcherQueue = matchers.get(regexp);
if (matcherQueue == null) {
synchronized (matchers) {
matcherQueue = matchers.get(regexp);
if (matcherQueue == null) {
// Create a new thread-safe Queue and a new Matcher
matcherQueue = new ConcurrentLinkedQueue<Matcher>();
matchers.put(regexp, matcherQueue);
} // Else: another thread already did what needed to be done
}
}
// Try to retrieve a Matcher
Matcher matcher = matcherQueue.poll();
if (matcher == null) {
// No matchers available, create one
// No lock needed, as it's not a drama to have a few more matchers in the queue
Pattern pattern = Pattern.compile(regexp);
matcher = pattern.matcher(value);
matcherQueue.offer(matcher);
} else {
// reset the matcher
matcher.reset(value);
}
// boolean notADuplicate = duplicateHunter.add(matcher.hashCode());
// if(!notADuplicate) {
// throw new RuntimeException("DUPLICATE!!!");
// }
return matcher;
}
private static void returnMatcher(String regex, Matcher matcher) {
Queue<Matcher> matcherQueue = matchers.get(regex);
//duplicateHunter.remove(matcher.hashCode());
matcherQueue.offer(matcher);
}
public static void main(String[] args) {
for (int i = 0; i < 2; i++) {
Thread thread = new Thread(new Runnable() {
public void run() {
for (int i = 0; i < 50000; i++) {
String regex = ".*";
Matcher matcher = null;
try {
matcher = getMatcher(regex, "toto" + i);
if (matcher.matches()) {
// matches
}
} finally {
if (matcher != null) {
returnMatcher(regex, matcher);
}
}
}
}
});
thread.start();
}
}
}
You'll get a "java.lang.StringIndexOutOfBoundsException: String index out of range". Enable the duplicateHunter code and you'll see that a Matcher is indeed returned twice sometimes.
(The static utility method isn't shown, the main method was made to show you the problem)
A:
When there are no matchers for a regexp, you create a new matcher, but you also add it to the queue right away:
if (matcher == null) {
// No matchers available, create one
// No lock needed, as it's not a drama to have a few more matchers in the queue
Pattern pattern = Pattern.compile(regexp);
matcher = pattern.matcher(value);
matcherQueue.offer(matcher); // Don't add it to the queue here!
}
Thus it will be in the queue while you are using it, and another thread could easily get a hold of it before you are done.
I don't know if I agree with your idea of pooling matchers by the way. They are not very expensive to create in terms of CPU cycles. You probably want to profile it to see if it's worth it. Precompiling the Pattern is a good idea, however.
|
堀江蘭最新番号
【KTDVR-147】熟女レイプ事件 10人4時間SP
【KT-321】美人人妻4人のレイプ事件9</a>2005-01-22KTファクトリー$$$SuperSLAV87分钟
|
var helper = require('./specHelper');
var tls = require('tls');
var fs = require('fs');
describe('tls', function(){
var serverKey;
var serverCert;
beforeEach( function() {
helper.testComplete(false);
serverKey = fs.readFileSync( './keys/RSA/server-key.pem');
serverCert = fs.readFileSync( './keys/RSA/server-cert.pem');
clientKey = fs.readFileSync( './keys/RSA/client-key.pem' );
clientCert = fs.readFileSync( './keys/RSA/client-cert.pem' );
})
it( 'should allow creation of a server with key and cert', function() {
var server = tls.createServer( {
key: serverKey,
passphrase: 'iamserver',
cert: serverCert,
},
function(connection) {
})
});
it( 'should allow creation of a server that can listen with key and cert', function() {
waitsFor(helper.testComplete, "server to be listening", 5000);
var server = tls.createServer( {
key: serverKey,
passphrase: 'iamserver',
cert: serverCert,
},
function(connection) {
})
server.listen(8181, function() {
server.close();
helper.testComplete(true);
});
});
it ('should allow a secure client connection', function() {
waitsFor(helper.testComplete, "server to receive connection", 5000);
var server = tls.createServer( {
key: serverKey,
passphrase: 'iamserver',
cert: serverCert
}, function(connection) {
expect( connection.getPeerCertificate() ).toBe( null );
connection.on( 'data', function(b) {
expect( b.toString() ).toBe( "howdy" );
connection.destroy();
server.close( function() {
helper.testComplete(true);
});
})
});
server.on('clientError', function(err) {
console.log( 'client error: ' + err );
})
server.listen( 8181, function() {
var client = tls.connect( 8181, { ca: [ serverCert ] }, function() {
expect( client.getPeerCertificate() ).not.toBe( null );
client.write( "howdy" );
})
})
});
it ('should allow a secure authenticated client connection', function() {
waitsFor(helper.testComplete, "server to receive connection", 5000);
var server = tls.createServer( {
key: serverKey,
passphrase: 'iamserver',
cert: serverCert,
requestCert: true,
rejectUnauthorized: true,
ca: [ clientCert ]
}, function(connection) {
expect( connection.getPeerCertificate() ).not.toBe( null );
connection.on( 'data', function(b) {
expect( b.toString() ).toBe( "howdy" );
connection.destroy();
server.close( function() {
helper.testComplete(true);
});
})
});
server.on('clientError', function(err) {
console.log( 'client error: ' + err );
})
server.listen( 8181, function() {
var client = tls.connect( 8181, {
ca: [ serverCert ],
key: clientKey,
passphrase: 'iamclient',
cert: clientCert
}, function() {
expect( client.getPeerCertificate() ).not.toBe( null );
client.write( "howdy" );
})
})
});
});
|
[Not Available].
The Minimum Basic Data Set (MBDS) is the most used database in the National Health System. The objective of this study is to evaluate the quality, in terms of coding errors of the administrative variables of the MBDS in the Basque Country Health Service. A retrospective, descriptive study by means of reviewing 768 clinical histories from 11 hospitals. the year 2005. The data obtained in the clinical histories were considered comparison standard and the level of discrepancy with the MBDS was established. The administrative variable measured were: sex, dates of birth, admission, discharge and of the first intervention, admission type and circumstances at discharge, department and physician responsible at discharge. The χ(2) was used to measure the between-hospital variability of the percentage discrepancy between the clinical histories and the MBDS records. There were 213,091 discharge reports analysed. The percentage errors in the MBDS varied between 0.7% in the sex variable and 24.9% in the physician responsible variable. The variability between hospitals was statistically significant (p=0.001). Good quality was observed in the records of the six variables: sex, date of birth, admission date, discharge date, admission type and medical department responsible. The records were lacking in the date of first intervention and the physician responsible variables.
|
117 F.3d 1433
17 O.S.H. Cas. (BNA) 2102
Amersig Southeastv.Sec. of Labor**
NO. 96-8993
United States Court of Appeals,Eleventh Circuit.
June 09, 1997
Appeal From: N.D.Ga. ,No.9301429OSHRC
1
Affirmed.
**
Local Rule 36 case
|
---
abstract: 'Polar codes are the first class of constructive channel codes achieving the symmetric capacity of the binary-input discrete memoryless channels. But the analysis and construction of polar codes involve the complex iterative-calculation. In this paper, by revisiting the error event of the polarized channel, a new concept, named polar spectrum, is introduced from the weight distribution of polar codes. Thus we establish a systematic framework in term of the polar spectrum to analyze and construct polar codes. By using polar spectrum, we derive the union bound and the union-Bhattacharyya (UB) bound of the error probability of polar codes and the upper/lower bound of the symmetric capacity of the polarized channel. The analysis based on the polar spectrum can intuitively interpret the performance of polar codes under successive cancellation (SC) decoding. Furthermore, we analyze the coding structure of polar codes and design an enumeration algorithm based on the MacWilliams identities to efficiently calculate the polar spectrum. In the end, two construction metrics named UB bound weight (UBW) and simplified UB bound weight (SUBW) respectively, are designed based on the UB bound and the polar spectrum. Not only are these two constructions simple and explicit for the practical polar coding, but they can also generate polar codes with similar (in SC decoding) or superior (in SC list decoding) performance over those based on the traditional methods.'
author:
- 'Kai Niu, , Yan Li, , Weiling Wu, [^1]'
title: 'Polar Codes: Analysis and Construction Based on Polar Spectrum'
---
Polar codes, Subcode, Polar subcode, Polar spectrum, Polar weight distribution, Union bound, Union-Bhattacharyya (UB) bound.
Introduction {#section_I}
============
Relative Research
-----------------
the first constructive capacity-achieving coding scheme, polar codes, invented by Ar[i]{}kan in 2009 are a great breakthrough [@Polarcode_Arikan] in channel coding theory. Since polar codes demonstrate advantages in error performance and other attractive application prospects, after nine years of research effort, polar codes became a coding standard for the control channel in the fifth generation (5G) wireless communication system [@5GNR_38212] in 2018. As the most significant concept, channel polarization was introduced to interpret the behaviour of polar coding. The Bhattacharyya parameter combined with the mutual information chain rule are used to evaluate the reliability of the polarized channel and to analyze the convergence behavior of polar codes. Consequently, this theoretical framework based on the channel polarization plays a key role to design and optimize the polar codes.
According to the dependence of the original channel information, the construction method of polar codes usually falls into two categories. In the first category, by using some channel parameters of the binary-input discrete memoryless channel (B-DMC), such as, erasure probability in binary erasure channel (BEC) or signal-to-noise ratio (SNR) in binary-input additive white Gaussian noise (BI-AWGN) channel, the reliability of the polarized channel can be iteratively calculated or evaluated based on the recursive structure of polar coding and the channel parameters of the original channels are not directly associated with the reliability. In what follows, this category is known as the channel-dependent or implicit construction. In the second category, the reliability of the polarized channel can be ordered based on some channel-independent characteristics of polar codes, which is named as the explicit construction. Due to the independence of channel condition, the explicit construction is more desired for the practical application.
For the implicit construction, Ar[i]{}kan first proposed a recursive calculation of Bhattacharyya parameter [@Polarcode_Arikan] to evaluate the reliability of the polarized channel, whereas this method is only precise for the coding construction in a BEC. For other B-DMCs, such as binary symmetry channel (BSC) or BI-AWGN channel, exact calculation of Bhattacharyya parameter will involve the high-complexity Monte-carlo integration while approximate iterative calculation will result in some performance losses. Subsequently, Mori *et al.* designed the density evolution (DE) algorithm to track the distribution of logarithmic likelihood ratio (LLR) and calculate the error probability under the successive cancellation (SC) decoding [@DE_Mori], which theoretically exhibits the highest accuracy. However, the high-precision DE algorithm has a complexity of $O(N\xi \log{\xi})$, where $N$ is the code length, $\xi$ denotes the number of samples and its typical value is about $10^5$, implying it’s a time-consuming process in practical application. In [@Tal_Vardy], Tal and Vardy proposed an iterative algorithm to evaluate the upper/lower bound on the error probability of each polarized channel, which can achieve almost the same accuracy as DE with a lower complexity of $O(N{\mu}^2\log\mu)$, where $\mu$ is a fixed integer far less than $\xi$. Afterwards, Trifonov advocated the use of Gaussian approximation (GA) to estimate the error probability in AWGN channel [@GA_Trifonov] thereby obtain good accuracy with further reduced complexity $O(N\log N)$. Lately, in order to further increase the accuracy of GA construction, we devised an improved GA algorithm [@IGA_Dai] for the polar codes with the long code length. Generally, all these algorithms belong to the implicit construction since the polarized channel reliability is derived based on the iterative calculation which relies on some parameters of the original channel and the large variation of channel condition may affect the reliability order of the polarized channel.
Commonly, from the viewpoint of system design, the coding construction should be explicit and independent of channel condition so as to facilitate the implementation of the encoder and decoder. Therefore, implicit construction is not convenient for the practical application of polar coding, on the contrary, the explicit construction is a more desirable selection. Sch$\rm\ddot{u}$rch *et al.* [@PO_Schurch] found that the reliability order of a part of polarized channels is invariant thereby introduced the concept of partial order (PO). It is an efficient tool to reduce the complexity of polar coding construction. Nevertheless, we still need to calculate the reliabilities of the rest polarized channels. Furthermore, He *et al.* [@PW_He] exploited the index feature of polarized channels and proposed an explicit construction, named polarized weight (PW) algorithm. Although PW is an empirical construction, amazingly, polar codes constructed by PW can achieve almost the same performance as those constructed by GA algorithm. In fact, the polar codes in 5G standard [@5GNR_38212] is constructed by using a fixed polar reliability sequence, which is universal for all the code configuration and obtained by computer searching [@5Gpolar_Bioglio].
Motivation
----------
Traditionally, good channel codes, such as Turbo/LDPC code, can be evaluated and optimized based on the distance spectrum or weight distribution [@Book_Lin]. However, the high complexity involved in the weight distribution calculation of polar codes [@Distance_spectrum; @Weight_distribution] makes it unrealistic to use such metrics for the polar code design. Although there are plenty of research results on the weight distribution in the classic coding theory, it is lack of the theoretic interpretation based on distance spectrum/weight distribution for the channel polarization. Obviously, there is a fault between the research of the classic channel coding and that of polar coding. Thus, in this paper, we focus on the theoretic framework based on distance spectrum to thoroughly interpret the behavior of polar codes and whereby deduce some new constructions that are analytical and explicit.
Main Contributions
------------------
In this paper, we introduce a new concept on the distance spectrum of polar codes, named polar spectrum, and establish a complete framework to analyze the performance of the polar codes under the SC decoding. Based on this framework, we obtain new explicit constructions for the polar coding. The main contributions of this paper can be summarized as follows.
1) First, the theoretical framework based on the polar spectrum is established. We revisit the error event analysis of polarized channel and find that one polarized channel is associated with a subcode and its codeword subset, namely polar subcode. Consequently, we set up a one-to-one mapping between the polarized channel and the polar subcode and introduce the weight distribution of the polar subcode, named polar spectrum, which is a set of the weight enumerator with the given Hamming weight. Compared with the Bhattacharyya parameter or other performance metrics needed iterative calculation, the union bound of the error probability of the polarized channel in term of polar spectrum has an intuitive interpretation, that is, the Hamming weight affects the pairwise error probability and the weight enumerator determines the number of the corresponding error events.
By the polar spectrum, we first build the relevance between the error probability of polarized channel and the weight distribution of polar code, which is a simple analytical-metric rather than a complex iterative-calculation. Furthermore, based on the polar spectrum, the union bound and the union-Bhattacharyya (UB) bound of the block error rate (BLER) under the SC decoding is derived. Meanwhile, the upper/lower bound of the mutual information of the polarized channel is also derived by using the polar spectrum. Therefore, the framework based on polar spectrum can provide the same performance analysis as the recursive calculation based on Bhattacharyya parameter.
2) Second, thanks to the good structure of (polar) subcodes, we design an iterative enumeration algorithm for the polar spectrum. We find that a pair of specific subcodes can constitute the dual codes. By using the well-known MacWilliams identities [@MacWilliams], the weight distribution of these two subcodes can be easily calculated. Moreover, we prove that the polar weight enumerator of polar subcode and the weight enumerator of subcode satisfy the recursive accumulation relation. On this basis, an iterative algorithm embedded the solution of MacWilliams identities is proposed to enumerate the polar spectrum of polar codes. Compared with the traditional searching algorithm of distance spectrum, such enumeration of polar spectrum is a low complexity and high efficiency algorithm, since its one running can generate all the polar spectra for arbitrary code configuration with a fixed code length.
3) Third, two explicit and analytical construction metrics, named union-Bhattacharyya weight (UBW) and simplified UB weight (SUBW), are proposed. They are the logarithmic version of the UB bound and the latter, SUBW, only considers the minimum weight term of the UB bound. Since the polar spectrum can be calculated off-line, these two constructions have a linear complexity far below those constructions based on the iterative calculation. On the other hand, they can also be modified as channel-independent constructions for the practical polar coding. Simulation results show that the polar codes constructed by these two metrics can achieve similar performance of those constructed by Tal-Vardy method, GA or PW algorithm under SC decoding. Dramatically, the former can outperform the latter under successive cancellation list (SCL) decoding.
The remainder of the paper is organized as follows. Section \[section\_II\] presents the preliminaries of polar codes, covering polar coding and decoding, error performance analysis and polar code construction. Section \[section\_III\] investigates the error event of polarized channel and introduces the concepts of subcode, polar subcode and polar spectrum. Then, by using polar spectrum, the union bound and union-Bhattacharyya bound of the error probability of the polarized channel are derived and analyzed. In addition, the mutual information of the polarized channel is also analyzed based on the polar spectrum. Furthermore, we explore the subcode duality of polar codes and design an iterative enumeration algorithm to calculate the polar spectrum in Section \[section\_V\]. Numerical analysis for the union bound and UB bound in term of polar spectrum and simulation results for comparing UBW/SUBW construction with the traditional constructions are presented in Section \[section\_VI\]. Finally, Section \[section\_VII\] concludes the paper.
Preliminary of Polar Codes {#section_II}
==========================
Notation Conventions
--------------------
In this paper, calligraphy letters, such as $\mathcal{X}$ and $\mathcal{Y}$, are mainly used to denote sets, and the cardinality of $\mathcal{X}$ is defined as $\left|\mathcal{X}\right|$. The Cartesian product of $\mathcal{X}$ and $\mathcal{Y}$ is written as $\mathcal{X}\times \mathcal{Y}$ and $\mathcal{X}^n$ denotes the $n$-th Cartesian power of $\mathcal{X}$. Especially, the hollow symbol, e.g. $\mathbb{C}$, denotes the codeword set of code or subcode. Let $\llbracket a, b \rrbracket$ denote the continuous integer set $\{a,a+1,\cdots,b\}$.
We write $v_1^N$ to denote an $N$-dimensional vector $\left(v_1,v_2,\cdots,v_N\right)$ and $v_i^j$ to denote a subvector $\left(v_i,v_{i+1},\cdots,v_{j-1},v_j\right)$ of $v_1^N$, $1\leq i,j \leq N$. Occasionally, we use the boldface lowercase letter, e.g. $\mathbf{u}$, to denote a vector. Further, given an index set $\mathcal{A}\subseteq \llbracket 1,N \rrbracket$ and its complement set $\mathcal{A}^c$, we write $v_\mathcal{A}$ and $v_{\mathcal{A}^c}$ to denote two complementary subvectors of $v_1^N$, which consist of $v_i$s with $i\in\mathcal{A}$ or $i\in\mathcal{A}^c$ respectively. Then $1_{\mathcal{A}}$ denotes the indicator function of a set $\mathcal{A}$, that is, $1_\mathcal{A}(x)$ equals $1$ if $x\in A$ and $0$ otherwise.
We use the boldface capital letter, such as $\mathbf{F}_N$, to denote a matrix with dimension $N$. So the notation $\mathbf{F}_N(a:N)$ indicates the submatrix consisting the rows from $a$ to $N$ of the matrix $\mathbf{F}_N$. We use $d_H(\mathbf{u},\mathbf{v})$ to denote the Hamming distance between the binary vector $\mathbf{u}$ and $\mathbf{v}$. Similarly, $w_H(\mathbf{u})$ denotes the Hamming weight of the binary vector $\mathbf{u}$. Given $\forall {\mathbf{a}}, {\mathbf{b}}\in \mathbb{R}^N$, let $\left\|\bf{a}-\bf{b}\right\|$ denote the Euclidian distance between the vector $\bf{a}$ and $\bf{b}$.
Throughout this paper, $\log\left(\cdot\right)$ means “logarithm to base 2,” and $\ln\left(\cdot\right)$ stands for the natural logarithm. $(\cdot)^T$ means the transpose operation of the vector. $(\mathbf{a}\cdot \mathbf{b})=\mathbf{a}\mathbf{b}^T$ is the inner product of two vectors $\mathbf{a}$ and $\mathbf{b}$. $\text{supp}(\mathbf{u})=\{i:u_i\neq0\}$ is the set of indices of non-zero elements of a vector $\mathbf{u}$. Let $\lceil x\rceil$ denote the ceiling function, that is, the least integer greater than or equal to $x$.
Encoding and Decoding of Polar Codes
------------------------------------
Given a B-DMC $W: \mathcal{X}\to \mathcal{Y}$ with input alphabet $\mathcal{X}= \{0,1\}$ and output alphabet $\mathcal{Y}$, the channel transition probabilities can be defined as $W(y|x)$, $x\in \mathcal{X}$ and $y\in \mathcal{Y}$. If $W$ is symmetric, there is a permutation $\pi_1$ on $\mathcal{Y}$ such that $W(y|1)=W(\pi_1(y)|0)$ for all $y\in \mathcal{Y}$. Let $\pi_0$ be the identity permutation on $\mathcal{Y}$. We write $x\star y$ to denote $\pi_x(y)$, here $x\in \mathcal{X}$ and $y\in \mathcal{Y}$.
Then the symmetric capacity and the Bhattacharyya parameter of the B-DMC $W$ can be defined as $$\label{equation1}
I(W) \triangleq \sum\limits_{y \in \mathcal{Y}} \sum\limits_{x \in \mathcal{X}} \frac{1}{2} W\left( y\left| x \right. \right)\log \frac{W( y\left| x \right.)}{\tfrac{1}{2}W(y\left| 0 \right.) + \tfrac{1}{2}W(y\left| 1 \right. )}$$ and $$\label{equation2}
Z( W ) \triangleq \sum\limits_{y \in \mathcal{Y}} {\sqrt {W( y|0)W(y|1)}}$$ respectively.
Applying channel polarization transform for $N=2^n$ independent uses of B-DMC $W$, after channel combining and splitting operation [@Polarcode_Arikan], we obtain a group of polarized channels $W_N^{(i)}: \mathcal{X}\to \mathcal{Y} \times \mathcal{X}^{i-1}$, $i\in \llbracket 1,N \rrbracket$. By using the channel polarization, the polar coding can be described as follows.
Given the code length $N$, the information length $K$ and code rate $R=K/N$, the indices set of polarized channels can be divided into two subsets: one set $\mathcal{A}$, named information set, to carry information bits and the other complement set $\mathcal{A}^c$ to assign the fixed binary sequence, named frozen bits. A message block of $K=|\mathcal{A}|$ bits is transmitted over the $K$ most reliable channels $W_N^{(i)}$ with indices $i\in \mathcal{A}$ and the others are used to transmit the frozen bits. So a binary source block $u_1^N$ consisting of $K$ information bits and $N-K$ frozen bits can be encoded into a codeword $x_1^N$ by $$\label{equation3}
x_1^N=u_1^N{\bf{F}}_N,$$ where the matrix ${{\bf{F}}_N}$ is the $N$-dimension generator matrix. [^2] This matrix can be recursively defined as ${{\bf{F}}_N} = {\bf{F}}_2^{ \otimes n}$, where “$^{\otimes n}$” denotes the $n$-th Kronecker product and ${{\bf{F}}_2} = \left[ { \begin{smallmatrix} 1 & 0 \\ 1 & 1 \end{smallmatrix} } \right]$ is the $2\times2$ kernel matrix.
Such polar coding can deduce the product-form channel $W^N$ and the synthetic channel $W_N$. The transition probabilities of these two channels satisfy $W_N\left( {y_1^N\left| {u_1^N} \right.} \right) = {W^N}\left( {y_1^N\left| {u_1^N{{\mathbf{F}}_N}} \right.} \right) = \prod\nolimits_{i = 1}^N {W\left( {{y_i}\left| {{x_i}} \right.} \right)}$. So the transition probability of the $i$-th polarized channel $W_N^{(i)}$ is defined as $$\label{equation4}
W_N^{(i)}\left( y_1^N,u_1^{i-1}\left| u_i \right. \right) \triangleq \sum\limits_{u_{i+1}^N \in {\mathcal{X}^{N-i}}} {\frac{1}{2^{N-1}} W_N\left( y_1^N\left| u_1^N \right. \right)}.$$ The Bhattacharyya parameter of $W_N^{(i)}$ is expressed as $$\label{equation5}
\begin{aligned}
&Z\left( {W_N^{\left( i \right)}} \right) \hfill \\
&= \sum\limits_{y_1^N \in {\mathcal{Y}^N}} {\sum\limits_{u_1^{i - 1} \in {\mathcal{X}^{i - 1}}} {\sqrt {W_N^{\left( i \right)}\left( {y_1^N,u_1^{i - 1}\left| 0 \right.} \right)W_N^{\left( i \right)}\left( {y_1^N,u_1^{i - 1}\left| 1 \right.} \right)} } }. \hfill \\
\end{aligned}$$
As proposed in [@Polarcode_Arikan], polar codes can be decoded by the SC decoding algorithm with a low complexity $O(N\log N)$. Furthermore, many improved SC decoding algorithms, such as successive cancellation list (SCL) [@SCL_Tal], successive cancellation stack (SCS) [@SCS_Niu], successive cancellation hybrid (SCH) [@SCH_Chen], successive cancellation priority (SCP) [@SCH_Guan], and CRC aided (CA)-SCL/SCS [@SCL_Tal; @CASCL_Niu; @ASCL_Li; @Survey_Niu] decoding can be applied to improve the performance of polar codes.
Error Performance of Polar Codes
--------------------------------
Based on the joint probabilities $P\left( {\left\{ {\left( {u_1^N,y_1^N} \right)} \right\}} \right) = {2^{ - N}}{W_N}\left( {y_1^N\left| {u_1^N} \right.} \right)$, we introduce the probability space $\left(\mathcal{X}^N \times \mathcal{Y}^N,P \right)$, where all $\left( {u_1^N,y_1^N} \right) \in \mathcal{X}^N \times \mathcal{Y}^N$. On this probability space, the random vectors $U_1^N$, $X_1^N$, $Y_1^N$, and $\hat {U}_1^N$ represent the input to the channel $W_N$, the input to the channel $W^N$, the output of $W^N$ or $W_N$, and the decisions by the decoder. Given a sample point $\left( {u_1^N,y_1^N} \right) \in \mathcal{X}^N \times \mathcal{Y}^N$, we have $U_1^N\left(u_1^N,y_1^N\right)=u_1^N$, $X_1^N\left(u_1^N,y_1^N\right)=u_1^N {\bf{F}}_N$, $Y_1^N\left(u_1^N,y_1^N\right)=y_1^N$, and $\hat {U}_1^N\left(u_1^N,y_1^N\right)$ is recursively determined by using SC decoding.
Given the fixed configuration $(N,K,\mathcal{A})$, the block error event under SC decoding is defined as $$\mathcal{E} \triangleq \left\{ {\left( {u_1^N,y_1^N} \right) \in {\mathcal{X}^N} \times {\mathcal{Y}^N}:{{\hat U}_\mathcal{A}}\left( {u_1^N,y_1^N} \right) \ne {u_\mathcal{A}}} \right\}.$$ So the block error rate can be expressed in this probability space as $P_e(N,K,\mathcal{A})=P(\mathcal{E})$. As the derivation in [@Polarcode_Arikan], the set of block error event can be enlarged as $\mathcal{E} \subset \mathop \bigcup \limits_{i \in \mathcal{A}} {\mathcal{E}_i}$, where the single-bit error event $\mathcal{E}_i$ is defined as $$\label{single_bit_error}
\begin{aligned}
\mathcal{E}_i \triangleq &\left\{\left(u_1^N,y_1^N\right)\in \mathcal{X}^N\times \mathcal{Y}^N: W_N^{(i)}\left(y_1^N,u_1^{i-1}\left|u_i\right.\right)\right.\\
&\left. \leq W_N^{(i)}\left(y_1^N,u_1^{i-1}\left|u_i\oplus 1\right.\right) \right\}.
\end{aligned}$$
Thus, the error probability of single-bit error event can be bounded as $P(\mathcal{E}_i)\leq Z\left(W_N^{(i)}\right)$. Further, the BLER under SC decoding can be upper bounded as $$\label{equation8}
P_e(N,K,\mathcal{A})=P(\mathcal{E})\leq \sum \limits_{i\in \mathcal{A}} P(\mathcal{E}_i) \leq \sum \limits_{i\in \mathcal{A}} Z\left(W_N^{(i)}\right).$$
Polar Code Construction
-----------------------
For the construction of polar codes, the calculation of channel reliabilities and selection of good channels are the critical steps. Generally, the construction algorithms can be divided into two categories, namely, the channel-dependent (implicit) construction and the explicit construction.
For the former, Ar[i]{}kan initially proposed the construction based on the Bhattacharyya parameter [@Polarcode_Arikan], whereas this method is only precise for the BEC and approximate for other channels. Lately, density evolution (DE) algorithm [@DE_Mori] and Tal-Vardy algorithm [@Tal_Vardy] were proposed to perform high-precise construction of polar codes. However, these algorithms have a slightly high complexity. The Gaussian approximation (GA) algorithm [@GA_Trifonov] is a desirable method to construct the polar codes with a medium complexity $O(N\log N)$, especially for the AWGN channel. Further, an improved GA algorithm [@IGA_Dai] was designed for the long code length. Nevertheless, these channel-dependent constructions are not convenient for the practical application.
On the contrary, since the explicit construction is independent with the channel condition, it is more desirable for the practical design of polar code. For an example, polarized weight (PW) construction [@PW_He] is a typical method in this category. Furthermore, due to the usage of constructive property of the generator matrix, the construction based on partial order [@PO_Schurch] is also a good method. However, these methods are still heuristic and do not fully explore the algebraic construction of polar code.
In this paper, we will re-establish the analysis framework of the error performance by investigating the distance spectrum of polar codes so as to obtain analytical construction metrics.
Performance Analysis based on Polar Spectrum {#section_III}
============================================
In this section, we will introduce a new analysis tool, namely polar spectrum, to analyze the error performance of polar codes under the SC decoding. First we investigate the error probability of single-bit error event and introduce the concepts of (polar) subcode and polar spectrum. Then we derive the new upper bounds of BLER based on polar spectrum. Further, as an enlarged version of BLER bound, the union-Bhattacharyya bound is also derived. Finally, the upper/lower bound of the mutual information of the polarized channel is analyzed by using this new tool.
Error Event Probability Analysis
--------------------------------
Recall that the single-bit error event $\mathcal{E}_i$ is associated with the $i$-th polarized channel, the probability of this error event can be upper bounded by summing those of many codeword error events.
\[theorem1\] Given $\forall u_{i+1}^N \in {\mathcal{X}^{N-i}}$ and suppose the transmission bit is zero, that is, $u_i=0$, then the error probability of the single-bit error event $\mathcal{E}_i$ can be enlarged as $$\begin{aligned}
P\left( {{{\cal E}_i}} \right) \le \sum\limits_{u_{i + 1}^N\in {{\cal X}^{N - i}}} &P\Big( {W_N}\left( {y_1^N\left| {0_1^N} \right.} \right) \\
& \le {W_N}\left( {y_1^N\left| {0_1^{i - 1},1,u_{i + 1}^N} \right.} \right) \Big).
\end{aligned}$$
By using Proposition 13 in [@Polarcode_Arikan], due to the symmetry of B-DMC $W$, the transition probabilities of synthetic channel $W_N$ and polarized channel $W_N^{(i)}$ can be written as $$\nonumber
\begin{aligned}
&W_N\left(\widetilde{y}_1^N\left|\widetilde{u}_1^{N}\right.\right)\\
&=W_N\left(a_1^N\mathbf{F}_N \star \widetilde{y}_1^N \left|\widetilde{u}_1^{N}\oplus a_1^{N}\right.\right)
\end{aligned}$$ and $$\nonumber
\begin{aligned}
&W_N^{(i)}\left(\widetilde{y}_1^N,\widetilde{u}_1^{i-1}\left|\widetilde{u}_i\right.\right)\\
&=W_N^{(i)}\left(a_1^N\mathbf{F}_N \star \widetilde{y}_1^N, \widetilde{u}_1^{i-1}\oplus a_1^{i-1}\left|\widetilde{u}_i\oplus a_i\right.\right)
\end{aligned}$$ respectively.
Let $a_1^{N}=\widetilde{u}_1^{N}$ and $a_1^N\mathbf{F}_N \star \widetilde{y}_1^N=y_1^N$, we have $$\nonumber
P\left(\mathcal{E}_i\right) =P\left(W_N^{(i)}\left(y_1^N,0_1^{i-1}\left|0\right.\right) \leq W_N^{(i)}\left(y_1^N,0_1^{i-1}\left| 1\right.\right) \right).$$
Extending the transition probabilities of polarized channel $W_N^{(i)}$ with Eq. (\[equation4\]), we have $$\nonumber
\begin{aligned}
P\left(\mathcal{E}_i\right)=
P& \left( \sum\limits_{\widetilde{u}_{i+1}^N \in {\mathcal{X}^{N-i}}} W_N\left( y_1^N\left| 0_1^{i-1},0,\widetilde{u}_{i+1}^N\oplus a_{i+1}^N \right. \right) \right.\\
&\left.\leq \sum\limits_{\widetilde{v}_{i+1}^N \in {\mathcal{X}^{N-i}}} W_N\left( y_1^N\left| 0_1^{i-1},1,\widetilde{v}_{i+1}^N \oplus a_{i+1}^N \right. \right) \right).
\end{aligned}$$ Let $\widetilde{v}_{i+1}^N \oplus a_{i+1}^N=u_{i+1}^N$, it follows that $$\begin{aligned}
P\left(\mathcal{E}_i\right)=
P&\left(\bigcup\limits_{u_{i+1}^N \in {\mathcal{X}^{N-i}}} \Bigg\{ W_N\left( y_1^N\left| 0_1^N \right. \right) \right.\\
&\leq W_N\left( y_1^N\left| 0_1^{i-1},1,u_{i+1}^N \right. \right) \Bigg\} \Bigg).
\end{aligned}$$ Using the property of the union set, we complete the proof.
Concerning the derivation of Theorem \[theorem1\], we find that the error event $\mathcal{E}_i$ is associated with the codewords transmitted over the polarized channel $W_N^{(i)}$. Thus we introduce the following definitions.
\[definition1\] Given the code length $N$, the $i$-th subcode $\mathbb{C}_N^{(i)}$ is defined as a set of codewords, that is, $$\mathbb{C}_N^{(i)}\triangleq \left\{{\bf{c}}:{\bf{c}}=\left(0_1^{(i-1)},u_i^N\right){\bf{F}}_N,\forall u_i^N \in \mathcal{X}^{N-i+1} \right\}.$$ Furthermore, one subset of the subcode $\mathbb{C}_N^{(i)}$, namely the polar subcode $\mathbb{D}_N^{(i)}$, can be defined as $$\mathbb{D}_N^{(i)}\triangleq \left\{{\bf{c}}^{(1)}:{\bf{c}}^{(1)}=\left(0_1^{(i-1)},1,u_{i+1}^N\right){\bf{F}}_N,\forall u_{i+1}^N \in \mathcal{X}^{N-i} \right\}.$$ The complement set of the polar subcode is defined as $\mathbb{E}_N^{(i)}=\mathbb{C}_N^{(i)}-\mathbb{D}_N^{(i)}\triangleq\left\{{\bf{c}}^{(0)}\right\}$.
Obviously, subcode $\mathbb{C}_N^{(i)}$ is a linear block code $(N,N-i+1)$ and we have $\left|\mathbb{C}_N^{(i)}\right|=2^{N-i+1}$ and $\left|\mathbb{D}_N^{(i)}\right|=2^{N-i}$. Let us investigate the Hamming distance between the codeword of $\mathbb{D}_N^{(i)}$ and that of $\mathbb{E}_N^{(i)}$.
\[proposition2\] For any ${\bf{c}}^{(1)}\in \mathbb{D}_N^{(i)}$ and ${\bf{c}}^{(0)}\in \mathbb{E}_N^{(i)}$, the Hamming distance between these two codewords satisfies $d_H\left({\bf{c}}^{(1)}, {\bf{c}}^{(0)}\right)=w_H\left({\bf{c}}^{(1)}\oplus {\bf{c}}^{(0)}\right)$ and the module-2 sum of these two codewords is belong to the given polar subcode, that is, ${\bf{c}}^{(1)}\oplus {\bf{c}}^{(0)} \in \mathbb{D}_N^{(i)}$.
By Definition \[definition1\], we have $$\nonumber
\begin{aligned}
&d_H\left({\bf{c}}^{(1)}, {\bf{c}}^{(0)}\right)\\
&=w_H\left(\left(\left(0_1^{(i-1)},1,u_{i+1}^N\right)\oplus \left(0_1^{(i-1)},0,v_{i+1}^N\right)\right) {\bf{F}}_N\right)\\
&=w_H\left(\left(0_1^{(i-1)},1,u_{i+1}^N\oplus v_{i+1}^N\right) {\bf{F}}_N\right).
\end{aligned}$$ Obviously, ${\bf{c}}^{(1)}\oplus {\bf{c}}^{(0)} \in \mathbb{D}_N^{(i)}$.
Based on Definition \[definition1\], we can further define the pairwise-codeword error event as follows.
\[definition2\] Given the codewords ${\bf{c}}^{(0)} \in \mathbb{E}_N^{(i)}$ and ${\bf{c}}^{(1)} \in \mathbb{D}_N^{(i)}$, the pairwise-codeword error event is defined as $$\begin{aligned}
\mathcal{D}_i\left({\bf{c}}^{(0)}\to {\bf{c}}^{(1)}\right)\triangleq &\left\{\left({\bf{c}}^{(0)},{\bf{c}}^{(1)}, y_1^N\right): W^N\left(y_1^N\left|{\bf{c}}^{(0)}\right.\right)\right. \\
&\left.\leq W^N\left(y_1^N\left|{\bf{c}}^{(1)}\right.\right)\right\}.
\end{aligned}$$
The error probability of the event $\mathcal{D}_i\left({\bf{c}}^{(0)}\to {\bf{c}}^{(1)}\right)$ is named as the pairwise error probability (PEP) and defined as $$\label{equation14}
\begin{aligned}
&P_N^{(i)}\left({\bf{c}}^{(0)}\to {\bf{c}}^{(1)}\right)=P\left( \mathcal{D}_i\left({\bf{c}}^{(0)}\to {\bf{c}}^{(1)}\right) \right)\\
&\triangleq \sum\limits_{{y_1^N}} W^N\left(y_1^N\left|\bf{c}^{(0)}\right.\right) 1_{\mathcal{D}_i\left({\bf{c}}^{(0)}\to {\bf{c}}^{(1)}\right)} \left({\bf{c}}^{(0)}, {\bf{c}}^{(1)},y_1^N\right).
\end{aligned}$$
From Proposition \[proposition2\], we can simplify the PEP as below, $$\label{equation18}
\begin{aligned}
&P_N^{(i)}\left({\bf{c}}^{(0)}\to {\bf{c}}^{(1)}\right)\\
&=P\left(W^N\left(y_1^N\left|{\bf{c}}^{(0)}\right.\right) \leq W^N\left(y_1^N\left|{\bf{c}}^{(1)}\right.\right)\right)\\
&=P\left(W^N\left(y_1^N\left|0_1^N \right.\right) \leq W^N\left(y_1^N\left|{\bf{c}}^{(0)}\oplus{\bf{c}}^{(1)}\right.\right)\right)\\
&=P_N^{(i)}\left(d_H\left({\bf{c}}^{(0)},{\bf{c}}^{(1)}\right)\right).
\end{aligned}$$ Without loss of generality, hereafter we designate ${\bf{c}}^{(0)}=0_1^N$. It follows that PEP $P_N^{(i)}\left({\bf{c}}^{(0)}\to {\bf{c}}^{(1)}\right)$ is determined by the codeword weight of polar subcode $\mathbb{D}_N^{(i)}$.
So we can derive the upper bound of the error probability of the polarized channel $W_N^{(i)}$ as follows.
\[theorem2\] The error probability of $W_N^{(i)}$ is upper bounded by $$P\left(W_N^{(i)}\right)\leq \sum\limits_{{\bf{c}}^{(1)}} P_N^{(i)}\left(d_H\left({\bf{c}}^{(0)},{\bf{c}}^{(1)}\right)\right).$$
According to Theorem \[theorem1\] and Definition \[definition2\], we have $$\label{equation17}
\begin{aligned}
P\left(W_N^{(i)}\right)&=P\left(\mathcal{E}_i\right)
\leq P\left(\bigcup \limits_{{\bf{c}}^{(1)}} \mathcal{D}_i\left({\bf{c}}^{(0)}\to {\bf{c}}^{(1)}\right)\right)\\
&= \sum \limits_{{\bf{c}}^{(1)}} P\left( \mathcal{D}_i\left({\bf{c}}^{(0)}\to {\bf{c}}^{(1)}\right) \right).
\end{aligned}\\$$ Substituting Eq. (\[equation18\]) into Eq. (\[equation17\]), we complete the proof.
Hereafter, $P_N^{(i)}\left(d_H\left({\bf{c}}^{(0)},{\bf{c}}^{(1)}\right)\right)$ is shortly written as $P_N^{(i)}(d)$.
From the above analysis, we establish the 1-1 mapping among the single-bit error event, the polarized channel, and the subcode, that is, $\left\{\mathcal{E}_i \leftrightarrow W_N^{(i)} \leftrightarrow {\mathbb{C}}_N^{(i)}\right\}$. Next we will further analyze the block error rate based on the distance spectrum of polar subcode.
Block Error Rate Bound Based on Polar Spectrum
----------------------------------------------
\[definition3\] The polar spectrum of the polar subcode $\mathbb{D}_N^{(i)}$, also named as polar weight distribution, is defined as the weight distribution set $\left\{A_N^{(i)}(d)\right\},d\in \llbracket 1, N \rrbracket $, where $d$ is the Hamming weight of non-zero codeword and the polar weight enumerator $A_N^{(i)}(d)$ enumerates the codewords of weight $d$ for codebook $\mathbb{D}_N^{(i)}$.
\[proposition3\] The error probability of $W_N^{(i)}$ is further upper bounded by $$\label{equation19}
P\left(W_N^{(i)}\right)\leq \sum \limits_{d=1}^N A_N^{(i)}(d) P_N^{(i)}(d).$$
Collecting the codewords with the same weight of the polar subcode $\mathbb{D}_N^{(i)}$ and using Definition \[definition3\], we complete the proof.
Essentially, this bound is the union bound over the polar subcode. So we further establish the 1-1 mapping among the single-bit error event, the polarized channel, and the polar subcode, that is, $\left\{\mathcal{E}_i \leftrightarrow W_N^{(i)} \leftrightarrow {\mathbb{D}}_N^{(i)}\right\}$. Let $d_{min}^{(i)}$ denote the minimum Hamming distance of polar subcode ${\mathbb{D}}_N^{(i)}$. Thus, by using the PEP formula (\[equation18\]), we obtain a union bound of BLER under SC decoding.
\[theorem3\] Given the fixed configuration $(N,K,\mathcal{A})$, the block error probability of polar code is upper bounded by $$\label{union_bound}
P_e(N,K,\mathcal{A})\leq \sum\limits_{i\in \mathcal{A}} \sum \limits_{d=1}^N A_N^{(i)}(d) P_N^{(i)}(d).$$
Compared with the upper bound (\[equation8\]) proposed by Ar[i]{}kan, this union bound has an analytical form, which is mainly determined by the polar spectrum of the selected polar subcodes, that is, the polar weight enumerators and the PEPs dominated by Hamming weights. So this upper bound can reveal more constructive features of polar codes than the traditional bounds. More tighter upper bounds, such as tangential bound or tangential-sphere bound (See [@Sason] and references therein), can also be used to evaluate the error performance of polar codes under SC decoding. Nevertheless, these improved upper bounds involve complex calculation and are not convenient for the practical application of polar coding. Therefore, in this paper, we focus on the simple upper bounds, such as union bound and union-Bhattacharyya bound. Next, we will further discuss the PEP under various B-DMC channels, such as BEC, BSC and BI-AWGN channel.
### PEP and BLER in the BEC
Given the BEC $W:\mathcal{X}\to \mathcal{Y}$, $\mathcal{X}=\{0,1\}$ and $\mathcal{Y}=\{0,e,1\}$ (Here, $e$ denotes an erasure), with the transition probabilities $W(y|x)$ and the erasure probability $\epsilon$, we have $W(e|0)=W(e|1)=\epsilon$.
Let $\mathcal{F}=\{i:y_i=e\}$ denote the set of erasure-occurred indices and assume $|\mathcal{F}|=l$. Assuming ${\bf{c}}^{(0)}=0_1^N$ and $d=d_H({\bf{c}}^{(0)},{\bf{c}}^{(1)})$, we conclude that only the case of the set of erasure-occurred indices covering the support set of ${\bf{c}}^{(0)}+{\bf{c}}^{(1)}$ may result in an error, that is, $l\geq d$ and $\text{supp}\left\{{\bf{c}}^{(0)}+{\bf{c}}^{(1)}\right\}\subseteq\mathcal{F}$. Hence, the PEP in the BEC can be expressed as $$\begin{aligned}
P_{BEC}\left({\bf{c}}^{(0)}\to{\bf{c}}^{(1)}\right)&=\sum \limits_{l=d}^{N}\left( {\begin{array}{c} N-d \\ l-d \end{array}} \right)\epsilon^{(l-d)}(1-\epsilon)^{N-l}\epsilon^d\\
&=\epsilon^d.
\end{aligned}$$ Furthermore, by Proposition \[proposition3\], the union bound of $W_N^{(i)}$ is derived as $$\label{equation22}
P_{BEC}\left(W_N^{(i)}\right)\leq \sum \limits_{d=d_{min}^{(i)}}^N A_N^{(i)}(d)\epsilon^d,$$ where $d_{min}^{(i)}$ is the minimum Hamming distance of polar subcode $\mathbb{D}_N^{(i)}$.
Correspondingly, by Theorem \[theorem3\], the upper bound of BLER using SC decoding in the BEC can be written as $$\label{equation23}
P_{e,BEC}(N,K,\mathcal{A})\leq \sum\limits_{i\in \mathcal{A}} \sum \limits_{d=d_{min}^{(i)}}^N A_N^{(i)}(d) \epsilon^d.$$
### PEP and BLER in the BSC
Suppose the BSC $W:\mathcal{X}\to \mathcal{Y}$, $\mathcal{X}=\{0,1\}$ and $\mathcal{Y}=\{0,1\}$, with the transition probabilities $W(y|x)$ and the crossover error probability $\delta$, we have $W(1|0)=W(0|1)=\delta$.
Given the transmission codeword ${\bf{c}}^{(0)}=0_1^N$, the received vector can be written as $y_1^N={\bf{c}}^{(0)}+e_1^N$, where $e_1^N$ is the error vector. So the PEP is derived as below.
\[theorem4\] Assuming the decision vector is ${\bf{c}}^{(1)}$ and $d=d_H({\bf{c}}^{(0)},{\bf{c}}^{(1)})$, the PEP in the BSC can be expressed as $$%\begin{aligned}
%&P_{BSC}\left({\bf{c}}^{(0)}\to{\bf{c}}^{(1)}\right)\\
%&=\sum\limits_{l=\lceil d/2\rceil}^{N} \sum\limits_{m=\lceil d/2 \rceil}^{d}
%\left( {\begin{array}{c} N-d \\ l-m\end{array}} \right) \left( {\begin{array}{c} d \\ m\end{array}} \right) \delta^l(1-\delta)^{N-l}.
P_{BSC}\left({\bf{c}}^{(0)}\to{\bf{c}}^{(1)}\right)
= \sum\limits_{m=\lceil d/2 \rceil}^{d}
\left( {\begin{array}{c} d \\ m\end{array}} \right) \delta^m(1-\delta)^{d-m}.
%\end{aligned}$$
When the pairwise error occurs, we have $d_H\left(y_1^N,{\bf{c}}^{(0)}\right)>d_H\left(y_1^N,{\bf{c}}^{(1)}\right)$. This condition is equal to $w_H\left(y_1^N+{\bf{c}}^{(0)}\right)>w_H\left(y_1^N+{\bf{c}}^{(1)}\right)$. Then we have $w_H\left(e_1^N\right)>w_H\left(e_1^N+{\bf{c}}^{(0)}+{\bf{c}}^{(1)}\right)=d_H\left(e_1^N,{\bf{c}}^{(0)}+{\bf{c}}^{(1)}\right)$. It follows that an error will occur if more than half of the elements in the support set of the error vector $e_1^N$ overlap the support set of the codeword ${\bf{c}}^{(0)}+{\bf{c}}^{(1)}$, that is, suppose $\text{supp}(e_1^N)=\mathcal{G}_1$ and $\left|\mathcal{G}_1\right|=m \geq \lceil d/2\rceil$, we have $\mathcal{G}_1\subseteq\text{supp}\left({\bf{c}}^{(0)}+{\bf{c}}^{(1)}\right)$. Thus we can enumerate the number of set partition of $\mathcal{G}_1$ and complete the proof.
Furthermore, by Theorem \[theorem4\], the union bound of $W_N^{(i)}$ is derived as $$%\begin{aligned}
P_{BSC}\left(W_N^{(i)}\right)\leq \sum \limits_{d=d_{min}^{(i)}}^N \sum\limits_{m=\lceil d/2 \rceil}^{d}\left( {\begin{array}{c} d \\ m\end{array}} \right) \delta^m(1-\delta)^{d-m}.
%\end{aligned}$$ Similarly, by Theorem \[theorem3\], the upper bound of BLER using SC decoding in the BSC can be written as $$\label{equation26}
%\begin{aligned}
P_{e,BSC}(N,K,\mathcal{A})\leq \sum\limits_{i\in \mathcal{A}} \sum \limits_{d=d_{min}^{(i)}}^N \sum\limits_{m=\lceil d/2 \rceil}^{d}\left( {\begin{array}{c} d \\ m\end{array}} \right) \delta^m(1-\delta)^{d-m}.
%\end{aligned}$$
### PEP and BLER in the AWGN channel
For a binary-input AWGN channel, the received signal is expressed as $$y_j=s_j+n_j,$$ where $s_j\in \{\pm \sqrt{E_s}\}$ is the BPSK signal, $E_s$ is the signal energy, and $n_j\sim\mathcal{N}(0,\frac{N_0}{2})$ is a Gaussian noise sample with the zero mean and the variance $\frac{N_0}{2}$. Since the BPSK modulation is used, the codeword ${\bf{c}}^{(0)}$ is transformed into the transmitted signal vector ${\bf{s}}^{(0)}=\sqrt{E_s}\left({\bf{1}}-2{\bf{c}}^{(0)}\right)$. Thus, the received signal vector can be addressed as $$\label{equation28}
{\bf{y}}=\sqrt{E_s}\left({\bf{1}}-2{\bf{c}}^{(0)}\right)+{\bf{n}},$$ where ${\bf{1}}$ is an all-one vector and ${\bf{n}}$ is the AWGN noise vector. Then the transition probability of the product-form channel can be written as $$W^N\left({\bf{y}\left|{\bf{c}}^{(0)}\right.}\right)=\frac{1}{\left(\pi N_0\right)^{N/2}}\exp\left\{ { - \frac{{{{\left\| {{\mathbf{y}} - {{\mathbf{s}}^{\left( 0 \right)}}} \right\|}^2}}}{{{N_0}}}} \right\}.$$ For the PEP of SC decoding in the AWGN channel, we have the following theorem.
\[theorem5\] Assuming the decision vector is ${\bf{c}}^{(1)}$ and $d=d_H\left({\bf{c}}^{(0)},{\bf{c}}^{(1)}\right)$, the PEP between ${\bf{c}}^{(0)}$ and ${\bf{c}}^{(1)}$ can be expressed as $$P_{AWGN}\left({\bf{c}}^{(0)}\to{\bf{c}}^{(1)}\right)=Q\left[\sqrt{\frac{2E_s}{N_0} d_H\left({\bf{c}}^{(0)},{\bf{c}}^{(1)}\right)}\right],$$ where $\frac{E_s}{N_0}$ is the symbol signal-to-noise ratio (SNR) and $Q(x)=\frac{1}{\sqrt{2\pi}}\int_x^{\infty} e^{-t^2/2}dt$ is the tail distribution function of the standard normal distribution.
First, we assume the codeword ${\bf{c}}^{(0)}$ and ${\bf{c}}^{(1)}$ are mapped to the transmission vector ${\bf{s}}^{(0)}$ and the signal vector ${\bf{s}}^{(1)}$ respectively. If a pairwise error occurs, the Euclidian distances among the received vector and the transmission/signal vectors satisfy the inequality ${\left\| {\mathbf{y}}- {\mathbf{s}}^{(0)} \right\|^2} > \left\| {\mathbf{y}}-{\mathbf{s}}^{(1)} \right\|^2$. Substituting (\[equation28\]), we have ${\left\| \mathbf{n}\right\|^2} > \left\|\left({\mathbf{s}}^{(0)}-{\mathbf{s}}^{(1)}\right)+{\mathbf{n}} \right\|^2$. Extending the inequality, we obtain the decision region $\mathcal{H}=\left\{\mathbf{n}:{\mathbf{n}}\left({\mathbf{s}}^{(0)}-{\mathbf{s}}^{(1)}\right)^T<- \frac{1}{2}\left\|{\mathbf{s}}^{(0)}-{\mathbf{s}}^{(1)}\right\|^2\right\}$. So the PEP can be written as $$\begin{aligned}
P\left({\mathbf{c}}^{(0)}\to{\mathbf{c}}^{(1)}\right)&= {\int { \cdots \int } }_{\mathcal{H}} {W_N}\left( {{\mathbf{y}}\left| {{{\mathbf{s}}^{\left( 0 \right)}}} \right.} \right)d{\mathbf{y}} \\
&= Q\left[ {\sqrt {\frac{{{1}}}{{2{N_0}}}} \left\| {{{\mathbf{s}}^{\left( 0 \right)}} - {{\mathbf{s}}^{\left( 1 \right)}}} \right\|} \right]. \\
\end{aligned}$$ Furthermore, due to $$\nonumber
\begin{aligned}
\left\|{\mathbf{s}}^{(0)}-{\mathbf{s}}^{(1)}\right\|&=\left\|\sqrt{E_s}\left[{\mathbf{1}}-2{\mathbf{c}}^{(0)}\right]-\sqrt{E_s}\left[{\mathbf{1}}-2{\mathbf{c}}^{(1)}\right]\right\|\\
&=2\sqrt{E_s}\left\|{\mathbf{c}}^{(1)}-{\mathbf{c}}^{(0)}\right\|,
\end{aligned}$$ then $\left\|{\mathbf{s}}^{(0)}-{\mathbf{s}}^{(1)}\right\|=2\sqrt{E_s d_H\left({\mathbf{c}}^{(0)},{\mathbf{c}}^{(1)}\right)}$. So we complete the proof.
Furthermore, by Theorem \[theorem5\], the union bound of $W_N^{(i)}$ is derived as $$P_{AWGN}\left(W_N^{(i)}\right)\leq \sum \limits_{d=d_{min}^{(i)}}^N A_N^{(i)}(d) Q\left(\sqrt{\frac{2dE_s}{N_0}}\right).$$ Similarly, by Theorem \[theorem3\], the upper bound of BLER using SC decoding in the BI-AWGN channel can be written as $$\label{equation33}
P_{e,AWGN}(N,K,\mathcal{A})\leq \sum\limits_{i\in \mathcal{A}} \sum \limits_{d=d_{min}^{(i)}}^N A_N^{(i)}(d) Q\left(\sqrt{\frac{2dE_s}{N_0}}\right).$$
Union-Bhattacharyya Bound
-------------------------
Union-Bhattacharyya (UB) bound provides a simple form to analyze the error performance bound of the B-DMC channel. Although the UB bound is slightly looser than the union bound, it is convenient to the theoretic analysis.
\[proposition4\] Given the B-DMC $W$, the union-Bhattacharyya (UB) bound of the polarized channel $W_N^{(i)}$ is given by $$\label{equation34}
P\left(W_N^{(i)}\right)\leq \sum \limits_{d=1}^N A_N^{(i)}(d) (Z(W))^d,$$ where $Z(W)$ is the Bhattacharyya parameter. Specifically, for the BEC, BSC and AWGN channel, the Bhattacharyya parameter $Z(W)$ is $\epsilon$, $2\sqrt{\delta(1-\delta)}$ and $\exp(-\frac{E_s}{N_0})$, respectively.
Furthermore, we can derive the UB bound of BLER as follows.
\[theorem6\] Given the fixed configuration $(N,K,\mathcal{A})$, the UB bound of the block error probability is written as $$\label{UB_bound}
P_e(N,K,\mathcal{A})\leq \sum\limits_{i\in \mathcal{A}} \sum \limits_{d=1}^N A_N^{(i)}(d) (Z(W))^d.$$
Following we will present the detailed form of UB bound in the BEC, BSC and AWGN channel.
Considering the Bhattacharyya parameter for the three channels, by Proposition \[proposition4\], the UB bound of the polarized channel $W_N^{(i)}$ is written as $$P\left( {W_N^{(i)}} \right) \le \left\{
{\begin{aligned}
&{\sum\limits_{d = d_{min}^{(i)}}^N {A_N^{(i)}} (d){\epsilon^d},{\rm{BEC}}}\\
&{\sum\limits_{d = d_{min}^{(i)}}^N {A_N^{(i)}} (d){{\left( {2\sqrt {\delta (1 - \delta )} } \right)}^d},{\rm{BSC}}}\\
&{\sum\limits_{d = d_{min}^{(i)}}^N {A_N^{(i)}} (d)\exp \left( { - \frac{{d{E_s}}}{{{N_0}}}} \right),{\rm{AWGN}}}
\end{aligned}} \right.$$ Note the UB bound is equivalent to the union bound (\[equation22\]) for the BEC. Similarly, according to Theorem \[theorem6\], given the fixed configuration $(N,K,\mathcal{A})$, the UB bound of the BLER using SC decoding in the BEC is equal to Eq. (\[equation23\]).
Further, the UB bounds of the BLER in the BSC and AWGN channel are respectively given by $$\label{equation38}
P_{e,BSC}(N,K,\mathcal{A})\leq \sum\limits_{i\in \mathcal{A}} \sum \limits_{d=d_{min}^{(i)}}^N A_N^{(i)}(d) {\left(2\sqrt{\delta(1-\delta)}\right)}^d,$$ and $$\label{equation40}
P_{e,AWGN}(N,K,\mathcal{A})\leq \sum\limits_{i\in \mathcal{A}} \sum \limits_{d=d_{min}^{(i)}}^N A_N^{(i)}(d) \exp\left(-\frac{dE_s}{N_0}\right).$$
Observing the union bound and UB bound of the polarized channel, we find that the channel parameters (erasure probability, crossover error probability or symbol SNR) and the weight distribution of the polar subcode explicitly determine the corresponding reliability. On the contrary, the reliability evaluation based on the Bhattacharyya parameter or other metrics needs an iterative calculation and cannot give a simple and direct formula. Since the union/UB bounds reveal the analytical form, we can easily use them to analyze and design the polar codes.
Mutual Information Analysis of Polarized Channel
------------------------------------------------
Based on the polar spectrum and PEP, we further analyze the symmetric capacity of the polarized channel $W_N^{(i)}$, that is, $I\left(W_N^{(i)}\right)=I\left(U_i;Y_1^N\left|U_1^{i-1}\right.\right)$. At first, we can derive the upper and lower bounds of the Bhattacharyya parameter $Z\left(W_N^{(i)}\right)$ as below.
\[proposition5\] Given the polarized channel $W_N^{(i)}$, the upper bound of the corresponding Bhattacharyya parameter satisfy the following inequality, $$\label{upper_bound_Z}
Z\left(W_N^{(i)}\right) \leq \sum \limits_{d=d_{min}^{(i)}}^N A_N^{(i)}(d) (Z(W))^d.$$
According to the definition in (\[equation5\]), the Bhattacharyya parameter can be further enlarged as $$\begin{aligned}
&Z\left( {W_N^{\left( i \right)}} \right) \\
&\begin{aligned}
\leq\frac{1}{2^{N-i}} \sum\limits_{y_1^N }&\sqrt {\sum\limits_{u_{i+1}^N }W_N\left( {y_1^N\left| 0_1^{i-1},0,u_{i+1}^{N} \right.} \right)} \\
&\cdot\sqrt{\sum\limits_{v_{i+1}^N }W_N\left( {y_1^N\left| 0_1^{i-1},1,v_{i+1}^N \right.} \right)}
\end{aligned} \\
&\begin{aligned}
\leq\frac{1}{2^{N-i}} \sum\limits_{y_1^N }\sum\limits_{u_{i+1}^N }\sum\limits_{v_{i+1}^N }&\sqrt {W_N\left( {y_1^N\left| 0_1^{i-1},0,u_{i+1}^{N} \right.} \right)}\\
&\cdot\sqrt{ W_N\left( {y_1^N\left| 0_1^{i-1},1,v_{i+1}^N \right.} \right)} .
\end{aligned}\\
\end{aligned}$$ Without loss of generality, let ${\bf{c}}^{(0)}=\left(0_1^{(i-1)},0,u_{i+1}^N\right){\bf{F}}_N=0_1^N$ and ${\bf{c}}^{(1)}=\left(0_1^{(i-1)},1,v_{i+1}^N\right){\bf{F}}_N$, by using the symmetry of subcode $\mathbb{C}_N^{(i)}$, we have $$\label{equation44}
Z\left( {W_N^{\left( i \right)}} \right) \leq \sum\limits_{y_1^N \in {\mathcal{Y}^N}}\sum\limits_{{\bf{c}}^{(1)}} {\sqrt {W^N\left( {y_1^N\left| {\bf{c}}^{(0)} \right.} \right) W^N\left( {y_1^N\left| {\bf{c}}^{(1)} \right.} \right)} }.$$ Enumerating all the codewords ${\bf{c}}^{(1)}$ of polar subcode $\mathbb{D}_N^{(i)}$, we obtain the upper bound of Bhattacharyya parameter.
If we only consider one codeword with the minimum Hamming weight $d_{min}^{(i)}$ in (\[upper\_bound\_Z\]), this term can be intended to serve as a lower bound to the Bhattacharyya parameter, that is, $$\label{Z_lower}
\left(Z(W)\right)^{d_{min}^{(i)}} \lesssim Z\left(W_N^{(i)}\right).$$
By using Proposition \[proposition5\], we can bound the symmetric capacity of any B-DMC $W$ as the following Theorem.
\[theorem8\] Given the polarized channel $W_N^{(i)}$ and the polar spectrum $\left\{A_N^{(i)}(d)\right\}$ of polar subcode $\mathbb{D}_N^{(i)}$, the symmetric capacity $I\left(W_N^{(i)}\right)=I\left(U_i;Y_1^N\left|U_1^{i-1}\right.\right)$ can be lower bounded by $$\label{I_lower}
%\begin{aligned}
I\left(W_N^{(i)}\right)\geq \max\left\{1-\log \left[1+\sum \limits_{d=d_{min}^{(i)}}^N A_N^{(i)}(d) (Z(W))^d\right],0\right\}.
%\end{aligned}$$
According to Proposition 1 in [@Polarcode_Arikan], the symmetric capacity $I\left(W_N^{(i)}\right)$ can be bounded by $$\label{I_upper_lower}
\begin{aligned}
\log \frac{2}{1+Z\left(W_N^{(i)}\right)}\leq I\left(W_N^{(i)}\right)\leq \sqrt{1-Z^2\left(W_N^{(i)}\right)}.
\end{aligned}$$ From Proposition \[proposition5\], we complete the proof.
Substituting (\[Z\_lower\]) into (\[I\_upper\_lower\]), the symmetric capacity $I\left(W_N^{(i)}\right)$ can be upper bounded by $$\label{I_upper}
I\left(W_N^{(i)}\right)\lesssim \sqrt{1-(Z(W))^{2d_{min}^{(i)}}}.$$
Especially, for the BEC, we can further derive the upper/lower bounds of the symmetric capacity of $W_N^{(i)}$ as below.
Given the BEC $W$ with the erasure probability $\epsilon$, then the symmetric capacity of the polarized channel $W_N^{(i)}$ satisfies $$\label{equation46}
I\left(W_N^{(i)}\right)\geq \max\left\{1-\sum \limits_{d=d_{min}^{(i)}}^N A_N^{(i)}(d) \epsilon^d,0\right\}.$$
Noting that $I\left(W_N^{(i)}\right)=1-P\left(W_N^{(i)}\right)\geq 0$ for the BEC and using Eq. (\[equation22\]), we obtain the lower bound.
Then by (\[I\_upper\]), it’s straightforward to derive the upper bound of $I(W_N^{(i)})$ for the BEC as $$\label{I_upper_BEC}
I\left(W_N^{(i)}\right)\lesssim \sqrt{1-\epsilon^{2d_{min}^{(i)}}}.$$
Fig. \[fig\_Channel\_polarization\] provides the upper/lower bounds of the symmetric capacity of the polarized channels for a BEC with the erasure probability $\epsilon=0.5$ and the code length $N=1024$. The symmetric capacity, marked by “I”, is evaluated by the recursive relation proposed by Ar[i]{}kan (See Eq. (38) in [@Polarcode_Arikan]). The upper and lower bounds, marked by “IUB” and “ILB” respectively, are calculated by (\[I\_upper\_BEC\]) and (\[equation46\]). Although the upper/lower bounds are the coarse approximation of the symmetric capacity, they also reveal the polarization phenomenon.
We establish an analytical framework to evaluate the error performance and the symmetric capacity of polar codes. The polar spectrum is the critical factor to indicate the bounds of the error probability and the mutual information of the polarized channel. So we can use the upper bound of the error probability as a metric to select the information set $\mathcal{A}$.
Calculation of Polar Spectrum {#section_V}
=============================
In this section, we will discuss the calculation of polar spectrum. First, we investigate the subcode duality of polar code. Then, by using such duality, we establish the MacWilliams identity between the subcodes. Finally, the enumeration algorithm of the polar spectrum is designed to calculate the polar weight enumerator.
Subcode Duality
---------------
Let $i \in \llbracket N/2+1, N\rrbracket$ denote the row index of the matrix $\mathbf{F}_N$. By the definition of subcode $\mathbb{C}_N^{(i)}$, its generator matrix $\mathbf{G}_{\mathbb{C}_N^{(i)}}$ is composed of the rows (from the $i$-th to the $N$-th row) of the matrix $\mathbf{F}_N$, that is, $\mathbf{G}_{\mathbb{C}_N^{(i)}}=\mathbf{F}_N(i:N)$. Further, we introduce another subcode $\mathbb{C}_N^{(N+2-i)}$, whose generator matrix satisfies $\mathbf{G}_{\mathbb{C}_N^{(N+2-i)}}=\mathbf{F}_N(N+2-i:N)$. Thus, subcode $\mathbb{C}_N^{(i)}$ is a linear code $(N,N-i+1)$ and its code rate is $R_{\mathbb{C}_N^{(i)}}=\frac{N-i+1}{N}=1-\frac{i-1}{N}$. Similarly, subcode $\mathbb{C}_N^{(N+2-i)}$ is a linear code $(N,i-1)$ and the code rate is $R_{\mathbb{C}_N^{(N+2-i)}}=\frac{i-1}{N}$.
\[theorem10\] Given $N/2+1\leq i\leq N$, subcode $\mathbb{C}_N^{(N+2-i)}$ is the dual code of subcode $\mathbb{C}_N^{(i)}$, that is, $\mathbb{C}_N^{(N+2-i)}=\mathbb{C}_N^{\bot(i)}$. Especially, $\mathbb{C}_N^{(N/2+1)}$ is a self-dual code.
We can use mathematical induction to prove this theorem. Given the code length $N$ and the index $N/2+1\leq i\leq N$, suppose the conclusion $\mathbb{C}_N^{(N+2-i)}=\mathbb{C}_N^{\bot(i)}$ holds. When the code length is doubled to $2N$ and the index is changed to $N+1\leq l\leq 2N$, we obtain two new subcodes $\mathbb{C}_{2N}^{(l)}$ and $\mathbb{C}_{2N}^{(2N+2-l)}$.
By the Plotkin structure $\left[\mathbf{u}+\mathbf{v} \left| \mathbf{v} \right.\right]$ of polar coding [@Polarcode_Arikan], due to $N+1\leq l\leq 2N$, the subcode $\mathbb{C}_{2N}^{(l)}$ consists of two identical component codes $\mathbb{C}_{N}^{(l-N)}$. That is to say, $\forall \mathbf{r}\in \mathbb{C}_{2N}^{(l)}$, we have $\mathbf{r}=(\mathbf{t},\mathbf{t}), \mathbf{t}\in \mathbb{C}_{N}^{(l-N)}$.
On the other hand, by the Poltkin structure, due to $2\leq 2N+2-l\leq N+1$, the subcode $\mathbb{C}_{2N}^{(2N+2-l)}$ is also regarded as a combination of two component codes: one is $\mathbb{C}_N^{(2N+2-l)}$ and the other is $\mathbb{C}_N^{(1)}$. That is, for every $\mathbf{w}\in \mathbb{C}_{2N}^{(2N+2-l)}$, it can be written as $\mathbf{w}=(\mathbf{a}+\mathbf{b},\mathbf{b})$, where $\mathbf{a}\in \mathbb{C}_N^{(2N+2-l)}$ and $\mathbf{b}\in \mathbb{C}_N^{(1)}$.
Now we calculate the inner product of $\mathbf{r}$ and $\mathbf{w}$, that is, $$\left({\mathbf{r}}\cdot {\mathbf{w}}\right) = \left({\mathbf{t}},{\mathbf{t}}\right)
\left( \begin{array}{*{20}{c}}
{\mathbf{a}}^T + {\mathbf{b}}^T \\
{\mathbf{b}}^T
\end{array}\right)
={\mathbf{t}}{\mathbf{a}}^T.$$ Due to ${\mathbf{t}}\in \mathbb{C}_N^{(i)}$ and ${\mathbf{a}}\in \mathbb{C}_N^{\bot(i)}$, we have ${\mathbf{t}}{\mathbf{a}}^T=0$. Thus, it follows that $\mathbb{C}_{2N}^{(2N+2-l)}=\mathbb{C}_{2N}^{\bot(l)}$.
Let $S_N^{(i)}(j)(1\leq j\leq N)$ denote the weight enumerators of the subcode $\mathbb{C}_N^{(i)}$, where $j$ is the Hamming weight of non-zero codeword of codebook $\mathbb{C}_N^{(i)}$. Similarly, $S_N^{\bot( i)}(j)$ denote the weight enumerators of the dual code $\mathbb{C}_N^{\bot (i)}$.
\[proposition8\] Given the subcode $\mathbb{C}_N^{(i)}$, we have $\mathbb{C}_N^{(i)}=\mathbb{D}_N^{(i)}\bigcup \mathbb{C}_N^{(i+1)}$. Thus, the weight enumerator and the polar weight enumerator satisfy $S_N^{(i)}(j)=A_N^{(i)}(j)+S_N^{(i+1)}(j)$.
Recall that the $i$-th subcode satisfies $\mathbb{C}_N^{(i)}=\mathbb{D}_N^{(i)}\bigcup \mathbb{E}_N^{(i)}$. Since $\mathbb{E}_N^{(i)}=\mathbb{C}_N^{(i+1)}$, then we have $\mathbb{C}_N^{(i)}=\mathbb{D}_N^{(i)}\bigcup \mathbb{C}_N^{(i+1)}$. It follows the relationship between the weight enumerator and the polar weight enumerator.
\[proposition9\] The odd codeword weight of the subcode $\mathbb{C}_N^{(i)}, i\in \llbracket 2,N \rrbracket$ is zero, that is, $S_N^{(i)}(2j+1)=0$. Similarly, for the polar subcode $\mathbb{D}_N^{(i)},i\in \llbracket 2,N \rrbracket$, we have $A_N^{(i)}(2j+1)=0$. For the polar subcode $\mathbb{D}_N^{(1)}$, the even codeword weight is zero, that is $A_N^{(1)}(2j)=0$.
Recall that ${{\bf{F}}_N} = {\bf{F}}_2^{ \otimes n}$, then we can obtain the conclusions by the following induction argument due to the Plotkin structure $[\bf{u}+\bf{v}|\bf{v}]$.
- Considering $n=1$, that is ${{\bf{F}}_2} = \left[ { \begin{smallmatrix} 1 & 0 \\ 1 & 1 \end{smallmatrix} } \right]$, it’s easy to obtain that $\begin{array}{*{20}{c}}
j&0&1&2\\
\hline
{S_2^{(1)}(j)}&1&2&1\\
\hline
{S_2^{(2)}(j)}&1&0&1
\end{array}$ and $\begin{array}{*{20}{c}}
j&0&1&2\\
\hline
{A_2^{(1)}(j)}&0&2&0\\
\hline
{A_2^{(2)}(j)}&0&0&1
\end{array}$. Obviously, the proposition holds under $n=1$.
- Assume the proposition holds under ${{\bf{F}}_N} = {\bf{F}}_2^{ \otimes n}$, where $n$ is an arbitrary integer that satisfies $n\geq2$, then consider the situation of ${{\bf{F}}_{2N}} = {\bf{F}}_2^{ \otimes (n+1)}$. Due to the matrix structure ${\mathbf{F}}_{2N}= \left[ { \begin{smallmatrix} {\mathbf{F}}_{N} & \mathbf{0} \\ {\mathbf{F}}_{N} & {\mathbf{F}}_{N} \end{smallmatrix} } \right]$, the discussion can be divided into three parts.
- First, in the case of $N+1\leq i\leq2N$, it’s straight to know that $S_{2N}^{(i)}(2j)=S_N^{(i-N)}(j)$ and $A_{2N}^{(i)}(2j)=A_N^{(i-N)}(j)$, where $j \in \llbracket 0, N\rrbracket$. Thus the odd weight enumerators of the subcode $\mathbb{C}_{2N}^{(i)}$ and polar subcode $\mathbb{D}_{2N}^{(i)}$ are zero.
- Second, in the case of $2\leq i\leq N$, consider $i=N$ first, for every codeword ${\bf{w}}\in\mathbb{C}_{2N}^{(N)}$, it can be written as ${\bf{w}}={\bf{a}}+{\bf{b}}$, where ${\bf{a}}=(\mathbf{t},\mathbf{0}), {\mathbf{t}}\in {\mathbb{C}}_{N}^{(N)}$ and ${\bf{b}} \in {\mathbb{C}}_{2N}^{(N+1)}$. From 2) and 3), it can be concluded that every ${\bf{w}}\in\mathbb{C}_{2N}^{(N)}$ has the even codeword weight since both ${\bf{a}}$ and ${\bf{b}}$ have the even weight. Next consider $i=N-1$, every codeword ${\bf{w}}\in\mathbb{C}_{2N}^{(N-1)}$ also can be written as ${\bf{w}}={\bf{a}}+{\bf{b}}$, where ${\bf{a}}=(\mathbf{t},\mathbf{0}), {\mathbf{t}}\in {\mathbb{C}}_{N}^{(N-1)}$ and ${\bf{b}} \in {\mathbb{C}}_{2N}^{(N)}$. From 2) and above analysis for $i=N$, it’s easy to know every ${\bf{w}}\in\mathbb{C}_{2N}^{(N-1)}$ has the even codeword weight. And so on, same conclusions can be drawn when $i=N-2,N-3,\cdots,2$. By using Proposition \[proposition8\], it can be also derived that $A_{2N}^{(i)}(2j+1)=0$.
- Finally, for $i=1$, every codeword ${\bf{w}}\in\mathbb{D}_{2N}^{(1)}$ can be written as ${\bf{w}}={\bf{c}}+{\bf{d}}$, where ${\bf{c}} = \left( {1,0,0, \cdots ,0} \right)$ is a $2N$-length vector with codeword weight one, and ${\bf{d}} \in {\mathbb{C}}_{2N}^{(2)}$ has the even weight from 4), whereby $\bf{w}$ has the odd codeword weight, i.e., $A_{2N}^{(1)}(2j)=0$. So, the proposition holds for ${{\bf{F}}_{2N}} = {\bf{F}}_2^{ \otimes (n+1)}$.
\[proposition10\] The weight distribution of the subcode $\mathbb{C}_N^{(i)}$ is symmetric, that is, $S_N^{(i)}(j)=S_N^{(i)}(N-j)$. Similar result holds for the polar subcode $\mathbb{D}_N^{(i)}$, that is, $A_N^{(i)}(j)=A_N^{(i)}(N-j)$.
Since all-one codeword is always contained in the set $\mathbb{C}_N^{(i)}$, we can easily obtain that $S_N^{(i)}(j)=S_N^{(i)}(N-j)$ by Theorem 1.4.5 (iii) in [@Book_Huffman]. Then by Proposition \[proposition8\], it can be also derived that $A_N^{(i)}(j)=A_N^{(i)}(N-j)$.
MacWilliams Identities of Subcodes
----------------------------------
It is well known that the linear relations between the weight distributions of a linear code and its dual can be determined by the MacWilliams identities [@MacWilliams]. That means, if we know the weight distribution of one linear code, we can obtain the weight distribution of the dual code without knowing the special coding structure. These identities have become the most significant tools to investigate and calculate the weight distributions.
\[theorem11\] Given the subcode $\mathbb{C}_N^{(i)}$ and its dual $\mathbb{C}_N^{\bot(i)}=\mathbb{C}_N^{(N+2-i)}$, the weight enumerators $S_N^{(i)}(j)$ and $S_N^{\bot(i)}(j)$ satisfy the following MacWilliams identities [@MacWilliams] $$\begin{aligned}
&\sum\limits_{j=0}^N \left(\begin{array}{*{20}{c}}N-j\\k\end{array}\right) S_N^{\bot(i)}(j)=2^{i-1-k} \sum\limits_{j=0}^N \left(\begin{array}{*{20}{c}}N-j\\N-k\end{array}\right)S_N^{(i)}{( j)},
\end{aligned}$$ where $k \in \llbracket 0,N\rrbracket$.
These identities are composed of $N+1$ linear equations. By solving these equations, we can calculate the weight distribution of one subcode and its dual.
Enumeration Algorithm of Polar Spectrum
---------------------------------------
Using the Plotkin structure $[\bf{u+v}|\bf{v}]$ and the MacWilliams identities, we design an iterative enumeration algorithm to calculate the polar spectrum. Given the weight distribution of subcodes $\mathbb{C}_N^{(i)}$, as shown in Algorithm \[algorithm1\], this algorithm enumerates the weight distribution of subcodes $\mathbb{C}_{2N}^{(i)}$ and the polar spectrum of polar subcodes $\mathbb{D}_{2N}^{(i)}$.
\[algorithm1\]
Based on the iterative structure, Algorithm \[algorithm1\] is consist of four steps to enumerate the weight distribution and the polar spectrum. In the first step, by using Proposition \[proposition8\], the polar spectrum of polar subcodes $\mathbb{D}_N^{(i)}$ with the code length $N$ is calculated. When the code length is grown from $N$ to $2N$, due to the matrix structure ${\mathbf{F}}_{2N}= \left[ { \begin{smallmatrix} {\mathbf{F}}_{N} & \mathbf{0} \\ {\mathbf{F}}_{N} & {\mathbf{F}}_{N} \end{smallmatrix} } \right]$, the enumeration can be divided into two parts.
That is to say, for the second step, we enumerate the weight distribution and polar spectrum in the case of $N+1 \leq l\leq 2N$. In this case, by using the Plotkin structure $\left[\mathbf{u}+\mathbf{v} \left| \mathbf{v} \right.\right]$, the subcode $\mathbb{C}_{2N}^{(l)}$ is consist of two identical component codes $\mathbb{C}_{N}^{(l-N)}$. Therefore, the subcode $\mathbb{C}_{2N}^{(l)}$ has the same weight distribution as the subcode $\mathbb{C}_{N}^{(l-N)}$. And similarly, the polar subcode $\mathbb{D}_{2N}^{(l)}$ has the same polar spectrum as the polar subcode $\mathbb{D}_{N}^{(l-N)}$. Certainly, the codeword weight of $\mathbb{C}_{2N}^{(l)}$ or $\mathbb{D}_{2N}^{(l)}$ is doubled.
For the third step, we enumerate the weight distribution and polar spectrum in the case of $1 \leq l\leq N$. According to Theorem \[theorem10\], the subcode $\mathbb{C}_{2N}^{(l)}$ is the dual of subcode $\mathbb{C}_{2N}^{(2N+2-l)}$. Since the weight distribution of $\mathbb{C}_{2N}^{(2N+2-l)}$ has been obtained in the second step, we can solve the MacWilliams identities to calculate the weight distribution of $\mathbb{C}_{2N}^{(l)}$.
Finally, using Proposition \[proposition8\], the polar spectrum of polar subcodes $\mathbb{D}_{2N}^{(l)}$ with the code length $2N$ is calculated. Note that the weight distribution of subcode $\mathbb{C}_{2N}^{(1)}$ obeys the binomial distribution.
The computational complexity of Algorithm \[algorithm1\] is mainly determined by the solution of MacWilliams identities. Due to the lower-triangle structure of the coefficient matrix in the MacWilliams identities, the weight enumerators can be recursively calculated. Hence, given the code length $N$, the worse complexity of solving MacWilliams identities is $\chi_M(N)=(N+1)^2$. Furthermore, we only consider to solve $N/2-1$ groups of those identities. So the computational complexity is $\chi_E(N)=(N/2-1)(N+1)^2$. If we consider the non-existence of the odd-weight codewords (Proposition \[proposition9\]) and the symmetry of the weight distribution (Proposition \[proposition10\]), the worse complexity of enumeration algorithm can be further reduced to $\chi_E(N)=(N/2-1)(N/4+1)^2$. So the total complexity of Algorithm \[algorithm1\] is $O(N^3)$.
Table \[polar\_spectrum\] shows partial results of the polar spectrum for the code length $N=32$. In this example, due to the symmetric distribution of polar spectrum (Proposition \[proposition10\]), we only provide half of the polar weight distribution (the number inside the bracket is the symmetric weight), e.g. $A_{32}^{(3)}(2)=A_{32}^{(3)}(30)=128$. As shown in this table, due to the duality relationship, $A_{32}^{(2)}$ and $A_{32}^{(32)}$, $A_{32}^{(3)}$ and $A_{32}^{(31)}$, etc. satisfy the MacWilliams identities. Meanwhile, the subcode $\mathbb{C}_{32}^{(17)}$ is a self-dual code. Furthermore, by Proposition \[proposition9\], all the polar weight distributions for the index $i\in \llbracket 2,32 \rrbracket$ have non-zero even codeword weight. On the contrary, for the index $i=1$, non-zero codeword weight is odd.
index $i$ weight $d$ $A_{32}^{(i)}(d)$ index $i$ weight $d$ $A_{32}^{(i)}(d)$
----------- ------------ ------------------- ----------- ------------ -------------------
1 1(31) 32 32 32 1
1 3(29) 4960 31 16 2
1 5(27) 201376 30 16 4
1 7(25) 3365856 29 8(24) 4
1 9(23) 28048800 28 16 16
1 11(21) 129024480 27 16 16
1 13(19) 347373600 27 8(24) 8
1 15(17) 565722720 26 16 32
2 2(30) 256 26 8(24) 16
2 4(28) 17920 25 12(20) 56
2 6(26) 453376 25 4(28) 8
2 8(24) 5258240 24 16 256
2 10(22) 32258304 23 16 192
2 12(20) 112892416 23 12(20) 128
2 14(18) 235723520 23 8(24) 32
2 16 300533760 22 16 384
3 2(30) 128 22 12(20) 256
3 4(28) 8960 22 8(24) 64
3 6(26) 226688 21 16 832
3 8(24) 2629120 21 12(20) 496
3 10(22) 16129152 21 8(24) 96
3 12(20) 56446208 21 4(28) 16
3 14(18) 117861760 20 16 1536
3 16 150266880 20 12(20) 1024
... ... ... 20 8(24) 256
17 2(30) 16 19 16 3200
17 6(26) 560 19 12(20) 2016
17 10(22) 4368 19 8(24) 448
17 14(18) 11440 19 4(28) 32
: Polar spectrum example for $N=32$[]{data-label="polar_spectrum"}
Construction of Polar Codes {#section_VI}
===========================
In this section, we consider the construction of polar codes based on the UB bound. First, we derive the logarithmic version metric based on UB bound, named UB weight (UBW). Then a more simple metric, named simplified UB weight (SUBW), is designed.
Construction Metrics based on UB bound
--------------------------------------
By Proposition \[proposition4\], we can use the UB bound of the channel error probability as a reliability metric to sort all the polarized channels. Consider the practical application, the logarithmic form of the UB bound is more convenient, which can be written as $$\begin{aligned}
&\ln \left\{\sum \limits_{d} A_N^{(i)}(d) (Z(W))^d\right\}\\
&=\ln \left\{\sum \limits_{d} \exp \left[\ln {A_N^{(i)}(d)}+d\ln(Z(W))\right]\right\}\\
&\approx \max_{d}\left\{L_N^{(i)}(d)+d\ln (Z(W))\right\}.
\end{aligned}$$ Here, we use the approximation $\ln\left( \sum\limits_{k} e^{a_k} \right)\approx \max\limits_{k}\{a_k\}$ and $L_N^{(i)}(d)=\ln\left[A_N^{(i)}(d)\right]$ is the logarithmic form of the polar weight enumerator. Then given the B-DMC $W$ and the code length $N$, the reliability of the polarized channel can be sorted by the [**[UB Weight]{}**]{} (UBW), that is, $$\label{UBW}
UBW_N^{(i)}=\max_{d}\left\{L_N^{(i)}(d)+d\ln (Z(W))\right\}.$$
(UBW in BEC or BSC) Given the Bhattacharyya parameter of the BEC $Z(W)=\epsilon$, UBW can be written as $$UBW_N^{(i)}=\max_{d}\left\{L_N^{(i)}(d)+d\ln \epsilon\right\}.$$ Similarly, UBW for the BSC $\left(Z(W)=2\sqrt {\delta(1-\delta)}\right)$ can be written as $$UBW_N^{(i)}=\max_{d}\left\{L_N^{(i)}(d)+\frac{d}{2}\ln [4\delta(1-\delta)]\right\}.$$
(UBW in AWGN channel) For the BI-AWGN channel $W$, the Bhattacharyya parameter is $Z(W)=\exp(-\frac{E_s}{N_0})$. Thus the corresponding UBW can be expressed as $$\label{UBW_AWGN}
UBW_N^{(i)}=\max_{d}\left\{L_N^{(i)}(d)-d\frac{E_s}{N_0}\right\}.$$ Especially, if the symbol SNR in UBW is instead of a fixed constant $\alpha$, we obtain a channel-independent construction metric.
Construction Metrics based on simplified UB bound
-------------------------------------------------
If we only consider the first term of the polar spectrum, that is, the minimum weight enumerator, the UBW given above can be further simplified as follows. Given the B-DMC $W$ and the code length $N$, the reliability of the polarized channel can be ordered by the [**[simplified UB Weight]{}**]{} (SUBW), that is, $${SUBW}_N^{(i)}=L_N^{(i)}\left(d_{min}^{(i)}\right)+d_{min}^{(i)}\ln (Z(W)).$$ Correspondingly, we can obtain the simplified union bound of the block error probability of polar codes $$P_e(N,K,\mathcal{A})\lesssim \sum\limits_{i\in \mathcal{A}} A_N^{(i)}\left(d_{min}^{(i)}\right) P_N^{(i)}\left(d_{min}^{(i)}\right)$$ or the simplified UB bound $$\label{equation65}
P_e(N,K,\mathcal{A})\lesssim \sum\limits_{i\in \mathcal{A}} A_N^{(i)}\left(d_{min}^{(i)}\right) (Z(W))^{d_{min}^{(i)}}.$$
(SUBW in AWGN channel) The SUBW for the BI-AWGN channel can be expressed as $$\label{SUBW_AWGN}
SUBW_N^{(i)}=L_N^{(i)}\left(d_{min}^{(i)}\right)-d_{min}^{(i)}\frac{E_s}{N_0}.$$ Similarly, we can design a channel-independent construction metric by selecting a suitable constant $\alpha$ for the symbol SNR in SUBW.
Table \[reliability\_order\] shows the reliability order of the polarized channels based on various constructions for the code length $N=32$ and the symbol SNR $\frac{E_s}{N_0}$ is fixed as $4$ dB. As shown in this table, the reliability order obtained by sorting UBW or SUBW from low to high is almost the same as that ordered by the GA algorithm except the polarized channels with indices 4 and 17. However, the reliability order obtained by sorting PW is different from that ordered by GA algorithm at the polarized channels with indices 20, 15, 8 and 25.
GA PW Bhattacharyya UBW SUBW
---- ---- --------------- ----- ------
32 32 32 32 32
31 31 31 31 31
30 30 30 30 30
28 28 28 28 28
24 24 24 24 24
16 16 16 16 16
29 29 29 29 29
27 27 27 27 27
26 26 26 26 26
23 23 23 23 23
22 22 22 22 22
20 15 20 20 20
15 20 15 15 15
14 14 14 14 14
12 12 12 12 12
8 25 8 8 8
25 8 25 25 25
21 21 21 21 21
19 19 19 19 19
13 13 13 13 13
18 18 18 18 18
11 11 11 11 11
10 10 10 10 10
7 7 7 7 7
6 6 6 6 6
4 4 4 17 17
17 17 17 4 4
9 9 9 9 9
5 5 5 5 5
3 3 3 3 3
2 2 2 2 2
1 1 1 1 1
The UBW and SUBW are deduced from the error probability of the polarized channel, that is, union-Bhattacharyya bound analysis. They are determined by the polar spectrum and the Bhattacharyya parameter. Compared with the traditional constructions, such as, DE, GA, and Tal-Vardy algorithm etc., these metrics have explicitly analytical structure. Particularly, since the polar spectrum can be off-line calculated by using Algorithm \[algorithm1\], the complexity of these constructions is linear $O(N)$, which is much lower than that of the former algorithms. On the other hand, compared with the SNR-independent polarized weight (PW) construction, these constructions can also be modified to the SNR-independent constructions. However, PW is an empirical construction and lack of theoretic explanation. On the contrary, UBW or SUBW has a good analytical property. Therefore, the construction based on UBW or SUBW has dual advantages: theoretic analyticity and practical application.
Numerical Analysis and Simulation Results {#section_VII}
=========================================
In this section, we will provide the numerical and simulation results based on the polar spectrum. First, we compare the upper bounds in term of the polar spectrum with the traditional bounds based on Tal-Vardy algorithm, GA or Bhattacharyya parameter under BEC, BSC and AWGN channel. Then, the BLER simulation results based on various constructions under AWGN channel are analyzed and compared.
Numerical analysis of upper bounds
----------------------------------
In this part, we compare six different upper bounds of BLER which can be divided into two categories, the traditional bounds and the proposed ones based on the polar spectrum. The traditional upper bounds include the Tal-Vardy bound [@Tal_Vardy], GA bound [@GA_Trifonov] and the bound provided by Ar[i]{}kan in [@Polarcode_Arikan] as presented in (\[equation8\]) while the proposed upper bounds contain the union bound given in (\[union\_bound\]), the UB bound expressed as (\[UB\_bound\]) and the simplified UB bound (\[equation65\]).
Given the code length $N = 128$, Fig. \[upper\_bound\_BEC\_BSC\] provides the upper bounds of BLER using SC decoding under BEC with the erasure probability $\epsilon=0.5$ and BSC with the crossover error probability $\delta=0.1$. The upper bound presented in (\[equation8\]) is named as “Ar[i]{}kan’s bound” hereafter and the lower bound marked by the dash line is obtained from $\max\limits_{i\in\mathcal{A}}\left\{Z\left(W_N^{(i)}\right)\right\}$ (See Fig. 7 in [@Polarcode_Arikan]). As shown in Fig. \[upper\_bound\_BEC\_BSC\](a), all the upper bounds grow larger as the code rate increases. However, it can be observed that the UB bound and the union bound (\[equation23\]) coincide under the BEC and they tend to diverge when the code rate is greater than $0.35$, while the simplified UB bound avoids this problem and is looser than the Ar[i]{}kan’s bound. When the BSC is considered, as shown in Fig. \[upper\_bound\_BEC\_BSC\](b), the union bound (\[equation26\]) is closer to the Ar[i]{}kan’s bound when the code rate is smaller than $0.3$, otherwise, the simplified UB bound is closer.
The BLER upper bounds of SC decoding under AWGN channel are depicted in Fig. \[upper\_bound\_AWGN\_N128\_1024\]. As shown in Fig. \[upper\_bound\_AWGN\_N128\_1024\](a), where the code length $N$ is set to $128$ and the code rate is $R=0.5$, all the upper bounds dramatically decrease with the increase of the symbol SNR. Due to the approximation calculation of Bhattacharyya parameter in AWGN channel, Ar[i]{}kan’s bound is looser than GA and Tal-Vardy bounds. In addition, the union bound (\[equation33\]) tends to be more closer to the simulation result (marked by dash line and constructed based on GA algorithm) than the Ar[i]{}kan’s bound when the symbol SNR exceeds $1.2$ dB. It also can be observed that the UB bound (\[equation40\]) and the simplified UB bound (\[equation65\]) are gradually close to the Ar[i]{}kan’s bound with the increase of the symbol SNR. Similar observations can be found in Fig. \[upper\_bound\_AWGN\_N128\_1024\](b), which provides the BLER upper bounds of SC decoding under AWGN channel with the configuration $N=1024$ and $R=0.5$.
Although the Tal-Vardy and GA bounds are tightly close to the BLER simulation result, they involve a complex on-line iterative-calculation. On the contrary, the proposed UB and simplified UB bounds not only have a linear complexity since the polar spectrum can be calculated off-line, but also can be used to deduce two explicit and analytical construction metrics to construct good polar codes. The BLER simulation results based on these explicit constructions will be presented in the next subsection.
Simulation Results
------------------
In this part, considering the AWGN channel, we compare the BLER simulation performances of polar codes generated by the proposed UBW/SUBW constructions and the traditional methods, which include Tal and Vardy’s algorithm [@Tal_Vardy], GA [@GA_Trifonov], PW [@PW_He] and the one based on Bhattacharyya parameter proposed by Ar[i]{}kan in [@Polarcode_Arikan]. The code length $N$ is in {128, 1024, 4096} and the code rate $R$ is choose from {$\frac{1}{3}$, $\frac{1}{2}$, $\frac{2}{3}$}. For Tal and Vardy’s method, the output alphabet size $\mu$ is set to $256$.
Fig. \[AWGN\_N128\_SC\_SCL\] provides the BLER performances comparison among the various constructions with $N = 128$ and $R=0.5$. For UBW/SUBW, the fixed symbol SNR in (\[UBW\_AWGN\])/(\[SUBW\_AWGN\]) is set to $\alpha=4$ dB. It can be observed from Fig. \[AWGN\_N128\_SC\_SCL\](a) that the polar codes constructed by UBW/SUBW can achieve similar performance of those constructed by Tal-Vardy, GA or PW algorithm under SC decoding. Moreover, the polar codes constructed by UBW/SUBW outperform those constructed based on the Bhattacharyya parameter in the case of $\frac{E_s}{N_0}>0.28 \text{dB}$. When the SCL decoding with list size $L=8$ is used, as shown in Fig. \[AWGN\_N128\_SC\_SCL\](b), the polar codes constructed by UBW/SUBW outperform those generated by the traditional constructions by about $0.4\sim0.6$ dB as the BLER is $10^{-3}$.
Given the code length $N=1024$, Fig. \[AWGN\_N1024\_SC\] presents the comprehensive BLER performances comparison among the various constructions under SC decoding. The symbol SNR in (\[UBW\_AWGN\]) for UBW is respectively set to $1.5$, $4$ and $4.5$ dB for the code rate $R=\frac{1}{3}$, $\frac{1}{2}$ and $\frac{2}{3}$, while that for SUBW in (\[SUBW\_AWGN\]) is respectively fixed as $1$, $3.5$ and $4$ dB. As shown in Fig. \[AWGN\_N1024\_SC\], when a low to medium code rate is considered, namely, $R=\frac{1}{3}$ or $\frac{1}{2}$, the polar codes constructed by UBW/SUBW perform slightly worse than those constructed by GA or Bhattacharyya parameter because the UB bound of the BLER under the SC decoding is looser than the GA bound or Ar[i]{}kan’s bound. However, similar to the PW construction, the UBW/SUBW constructions ensure the explicity at the cost of some performance losses. In addition, when the code rate $R=\frac{2}{3}$ is considered, the polar codes constructed by UBW/SUBW can achieve nearly the same performance of those constructed by Tal-Vardy or GA. Apart from these, one can also observe that the polar codes constructed by PW show error floor in the high SNR region, which can be avoided by UBW/SUBW.
For the code length $N=1024$, the BLER performances comparison among the various constructions under SCL decoding with list size $L=16$ is shown in Fig. \[AWGN\_N1024\_SCL\]. It can be observed that the polar codes constructed by UBW/SUBW can achieve better performance than those constructed by Tal-Vardy/GA/PW or the construction based on Bhattacharyya parameter. The performance gain of polar codes constructed by UBW/SUBW becomes larger with the increase of code rate. For example, given $R=\frac{1}{2}$, when the BLER is $10^{-3}$, polar codes constructed by SUBW achieve $0.3$/$0.5$ dB gain compared to those constructed by GA/PW, it then becomes $0.4$/$1.3$ dB under $R=\frac{2}{3}$. Actually, polar codes constructed by UBW/SUBW benefit from the adequate utilization of the polar spectrum, which is vital for the polar code construction under the SCL decoding. Furthermore, polar codes constructed by SUBW outperform those constructed by UBW when $R=\frac{1}{3}$ because the minimum weight term of the polar spectrum plays an important role in the analysis of error probability of the polarized channels.
Fig. \[AWGN\_N4096\_SC\_SCL\] provides the BLER performances comparison among the various constructions with $N = 4096$ and $R=\frac{2}{3}$. In this case, the symbol SNR in (\[UBW\_AWGN\])/(\[SUBW\_AWGN\]) for UBW/SUBW is set to $4.5$/$3.5$ dB. It can be observed from Fig. \[AWGN\_N4096\_SC\_SCL\](a) that polar codes constructed by UBW/SUBW perform worse than those constructed by Tal-Vardy or GA under SC decoding but avoid the error floor in the high SNR region, which is obviously shown in the polar codes constructed by PW. When the SCL decoding with list size $L=16$ is used, as shown in Fig. \[AWGN\_N4096\_SC\_SCL\](b), the polar codes constructed by SUBW can achieve $0.5$/$1.1$ dB gain compared to those constructed by Bhattacharyya parameter/PW when the BLER is $10^{-3}$.
In general, UBW and SUBW can be regarded as two good constructions, which are simple and explicit for the practical polar coding as well as able to generate polar codes with superior performance under SCL decoding over those based on the traditional methods in most cases.
Conclusions {#section_VIII}
===========
In this paper, we introduce a new concept named polar spectrum from the codeword weight distribution of polar codes and then establish a systematic framework in term of the polar spectrum to analyze and construct polar codes. On the basis of the polar spectrum, we derive the union bound and UB bound of the polarized channel and further upper bound the BLER of SC decoding. In addition, we propose an iterative algorithm embedded the solution of MacWilliams identities to enumerate the polar spectrum of polar codes, which has a low complexity and a high efficiency compared with the traditional searching algorithm. Finally, we design two explicit and analytical construction metrics named UBW and SUBW, which have a linear complexity far below those constructions based on the iterative calculation since the polar spectrum can be calculated off-line. Simulation results show that the polar codes constructed by these two metrics can achieve similar performance of those constructed by GA or PW algorithm under SC decoding and even superior performance under SCL decoding.
[99]{}
E. Ar[i]{}kan, “Channel polarization: a method for constructing capacity achieving codes for symmetric binary-input memoryless channels,” *IEEE Trans. Inf. Theory*, vol. 55, no. 7, pp. 3051-3073, July 2009.
S. B. Korada, “Polar Codes for channel and source coding,” Dissertation of EPFL, 2009.
E. Ar[i]{}kan and E. Telatar, “On the rate of channel polarization,” *in Proc. IEEE Int. Symp. Inform. Theory (ISIT)*, pp. 1493-1495, July 2009.
R. Mori and T. Tanaka, “Performance of polar codes with the construction using density evolution,” *IEEE Commun. Lett.*, vol. 13, no. 7, pp. 519-521, Jul. 2009.
I. Tal and A. Vardy, “How to construct polar codes,” *IEEE Trans. Inf. Theory*., vol. 59, no. 10, pp. 6562-6582, 2013.
P. Trifonov, “Efficient design and decoding of polar codes,” *IEEE Trans. Commun.*, vol. 60, no. 11, pp. 3221-3227, Nov. 2012.
Dai J. C., Niu K., et al., “Does Gaussian Approximation Work Well for the Long-Length Polar Code Construction?” IEEE Access, vol. 5, pp. 7950-7963, 2017.
C. Sch$\rm\ddot{u}$rch, “A partial order for the synthesized channels of a polar code,” *IEEE International Symposium on Information Theory (ISIT)*, pp. 220-224, July 2016.
G. N. He, J. C. Belfiore , et al., “$\beta$-expansion: A Theoretical Framework for Fast and Recursive Construction of Polar Codes,” *IEEE GLOBECOM*, pp. 1-6, Dec. 2017.
V. Bioglio, C. Condo, and I. Land, “Design of Polar Codes in 5G New Radio,” Arxiv 1804.04389v1, Apr. 2018.
M. Valipour and S. Yousefi, “On probablistic weight distribution of polar codes,” *IEEE Commun. Lett.*, vol. 17, no. 11, pp. 2120-2123, 2013.
Z. Z. Liu, K. Chen, K. Niu, and Z. Q. He, “Distance spectrum analysis of polar codes,” *in Proc. IEEE Wireless Commun. Networking Conf. (WCNC)* , pp. 490-495, April 2014.
I. Tal and A. Vardy, “List decoding of polar codes,” *in Proc. IEEE Int. Symp. Inform. Theory (ISIT)*, pp. 1-5, 2011.
K. Niu and K. Chen, “CRC-aided decoding of polar codes,” *IEEE Commun. Lett.*, vol. 16, no. 10, pp. 1668-1671, Oct. 2012.
B. Li, H. Shen, and D. Tse, “An adaptive successive cancellation list decoder for polar codes with cyclic redundancy check,” IEEE Commun. Lett., Vol. 16, No. 12, pp. 2044-2047, 2012.
K. Niu, K. Chen, J. R. Lin, and Q. T. Zhang, “Polar codes: primary concepts and practical decoding algorithms,” *IEEE Commun. Mag.*, pp. 192-203, July 2014.
K. Niu and K. Chen, “Stack decoding of polar codes,” *Electronics Letters*, vol. 48, no. 12, pp. 695-697, 2012.
K. Chen, K. Niu, and J. R. Lin, “Improved successive cancellation decoding of polar codes,” *IEEE Trans. Commun.*, vol. 61, no. 8, pp. 3100-3107, 2013.
D. Guan, K. Niu, C. Dong, and P. Zhang, “Successive cancellation priority decoding of polar codes,” *IEEE Access*, vol. 7, pp. 9575-9585, 2019.
F. J. MacWilliams, “A theorem on the distributionof weights in a systematic code,” *Bell System Tech.*, vol. 42, pp. 79-94, 1963.
I. Sason and S. Shamai, “Performance analysis of linear codes under maximum-likelihood decoding: a tutorial,” *Foundations and Trends $^\circledR$ in Communications and Information Theory*, vol. 3, no. 1/2, pp. 1-225, 2006.
S. Lin and D. J. Costello Jr., *Error Control Coding: Fundamentals and Applications (2nd ed.)*, Pearson Education, 2004.
W. C. Huffman and V. Pless, *Fundamentals of Error Correcting Codes*, Cambridge, U.K.: Cambridge Univ. Press, 2003. 3rd Generation Partnership Project (3GPP) TS 36.212, “Multiplexing and channel coding,” Release 8, 2009.
3rd Generation Partnership Project (3GPP) TS 38.212, “Multiplexing and channel coding,” V.15.1.0, 2018.
Qualcomm, “LDPC Rate Compatible Design Overview,” 3GPP TSG R1-1610137, Lisbon, Portugal, Oct. 2016.
[^1]: This work is supported by the National Key Research and Development Program of China (No. 2018YFE0205501), National Natural Science Foundation of China (No. 61671080) and Qualcomm Corporation. K. Niu, Y. Li, W. Wu are with the Key Laboratory of Universal Wireless Communications, Ministry of Education, Beijing University of Posts and Telecommunications, Beijing, 100876, China (e-mail: {niukai, Lyan, weilwu}@bupt.edu.cn).\
[^2]: In the seminal paper [@Polarcode_Arikan], the generator matrix is composed of the matrix ${\bf{F}}_N$ and the bit-reversal matrix. Since the bit-reversal operation does not affect the reliability of the polarized channel, in this paper, we will use the matrix $\mathbf{F}_N$ as the generator matrix. This is also the polar coding form in 5G standard [@5GNR_38212].
|
import absence from './absence'
import acceptance from './acceptance'
import addValidator from './add-validator'
import confirmation from './confirmation'
import date from './date'
import email from './email'
import file from './file'
import format from './format'
import inclusion from './inclusion'
import exclusion from './exclusion'
import length from './length'
import numericality from './numericality'
import presence from './presence'
import url from './url'
import combine from './combine'
import validateForm from './validate-form'
import Validators from './validators'
let required = presence // alias
export default Validators
export {
absence,
acceptance,
addValidator,
confirmation,
date,
email,
exclusion,
file,
format,
inclusion,
length,
numericality,
presence,
required,
url,
combine,
validateForm
}
|
This is an archived article and the information in the article may be outdated. Please look at the time stamp on the story to see when it was last updated.
Please enable Javascript to watch this video
CLAYTON, MO (KTVI) – Longtime conservative activist and Eagle Forum Founder Phyllis Schlafly believes her organization is in the middle of a legal fight because she is a Donald Trump supporter. It’s a claim disputed by attorneys who have filed a lawsuit in Madison County, IL representing several of the forum’s board members.
Schlafly said Tuesday, “The biggest issue right now is immigration.” She went on to say it’s time to remove so-called party kingmakers who have, “Given us a long series of losers.”
The lawsuit seeks to remove Schlafly’s hand-picked Eagle Forum president from his position, and to give board members greater authority. The forum’s board recently held a meeting to remove Ed Martin. The meeting was held by phone. Plaintiffs’ attorneys argued board members felt Martin exceeded his authority.
“Well I think it’s pretty clear that they want to take over Eagle Forum, and run it their way,” said Schlafly. “I built a very successful organization.”
Attorney Scott Clark represents the defendants. He said the vote to remove Martin is invalid. Clarks said the Eagle Forum’s bylaws require the board to meet in person.
“This is an attempt to steal the organization, the Eagle Forum organization from Mrs. Schlafly,” said Clark. “She’s worked more than 40 years or more to build the organization based on her reputation, her values, her hard work.”
Attorneys for the Plaintiffs declined to be interviewed. They argued in court that board members had the legal authority to meet by phone. They also argued board member motives were not based on party politics, rather the ability to carry out their duties.
The heart of the case centers on whether the board can meet by phone, or must meet in person. The judge hearing the case will first make a ruling on that issue, before deciding whether the case will go forward.
|
Arlene sacked? Time to step up Ruby, Zoe and Vanessa
To the picket lines, Maureen Lipman! Placard at the ready Esther Rantzen! Zoe Wanamaker, Ruby Wax, pick up your megaphones! You are who we need at this time of crisis. I’m proposing a new union, Older Jewish Entertaining Women (O JEW) to take on the bumbling bosses of the BBC. We muttered about Wossy and Sachsgate. We grumbled over the Blue Peter cat row. We rolled our eyes at executives’ vast expenses claims. But this is a step too far. Arlene Phillips has been booted off the Strictly Come Dancing judging panel.
You may fail to see the Jewish — or any — interest in a competition for two-left-footed celebrities, despite the fine performances of co-religionists like Natasha Kaplinsky and Rachel Stevens. Maybe you prefer the hora to the foxtrot.
If so, you have missed the star that is Arlene. Warm, wise, witty and slightly bonkers she is a composite of everyone’s favourite auntie from Prestwich.
She is a complete expert in her field, as erudite in her own way as Rabbi Julia Neuberger or Baroness Ruth Deech. When Arlene attacked John Sergeant’s lamentable footwork her comments were based upon years of choreographing the likes of Hot Gossip. Arlene knows her stuff —unlike Alesha Dixon, her replacement, picked to attract younger viewers.
Arlene’s judgements are packed with wordplay and delivered with aplomb. She knows how to send herself up, but takes no nonsense from high-spirited fellow judges. Like Ruby Wax or Vanessa Feltz she is confident enough to joke about her own sexuality — but knows where to draw the line.
She’s got the caustic edge of Wanamaker or Lipman, and the ability to wound of Joan Rivers. Like many older Jewish woman Arlene delivers outrageously insulting criticism cleverly disguised as helpful advice. “No one’s remarks cut as deeply as Arlene’s, yet backstage she was like a caring mother figure,” remarked flat-footed TV presenter Fiona (no relation) Phillips, survivor of many on-screen maulings from Arlene. Fiona’s confusion will be familiar to anyone whose booba has weighed in on a radical haircut or an ill-fitting yomtov outfit.
Last but not least Arlene, 66 looks stunning — and not a day older than 40. Just like most older Jewish woman. Look at Felicity Kendal. So Arlene’s departure is no small matter. Not only has the BBC broken every employment law about sexism and ageism, but they have failed to cherish a Jewish treasure. Solidarity is essential. Maureen, Zoe, Felicity, no more heavy drama or jolly sitcoms. Ruby, zip that mouth; Esther, Julia and Ruth, just say no to Question Time. Can someone get Madonna on board?
Before we unleash O JEW’s industrial action though, there are two ways that the BBC can avoid this blow.
Brucie’s getting on a bit. Let’s have Arlene as Strictly’s presenter. Of course, she’ll need some eye candy at her side. Sweet Gethin Jones or Matt di Angelo? That’ll pull in viewers, young and old.
But maybe Strictly is too small for Arlene. There’s a vacancy at another top BBC show. The Apprentice needs a new right-hand woman, now that Margaret Mountford is to spend more time with her PhD. Who better to take her place than sparky Arlene? Sir Alan… it’s up to you.
|
An improved method for the determination of potassium in leucocytes.
1. In twenty normal subjects and five patients with disturbances of potassium balance, the potassium concentration in leucocytes was measured by a modified technique. 2. The precision of the method was three to four times greater than that of techniques previously described, with a coefficient of variation for duplicate analyses of 2.3%. 3. A comparison between intraleucocyte potassium concentration and total body potassium in five patients with pathological alterations in their potassium balance showed that intracellular potassium concentration may reflect the clinical state better than total body potassium.
|
Photoactivatable substrates for systematic study of the impact of an extracellular matrix ligand on appearance of leader cells in collective cell migration.
Epithelial cells migrate as multicellular units. The directionality and speed of these units are determined by actively moving leader cells. It is important to understand how external cues affect the appearance of these leader cells in physiological and pathological processes. However, the impact of extracellular matrices (ECMs) is still controversial, because physically-adsorbed ECM proteins are amenable to protein remodeling, and uncontrolled cluster geometry can vary migration phenotypes. Here, we demonstrate a photoactivatable substrate, which we used to study the impact of a cyclic Arg-Gly-Asp (cRGD) ligand on leader cell formation in MDCK cells. This robust platform allowed us to investigate the effect of cRGD density on leader cell formation, in any given cluster geometry, with minimized ECM remodeling. Our results show a biphasic response of leader cell appearance upon reducing the surface cRGD density. The increase, in leader cell appearance, within the higher density range, is not only associated with the weakening of circumferential actomyosin belts, but also reduction of cellular mechanical tension and intercellular junctional E-cadherin. These results indicate that cRGD-mediated cell-ECM interactions positively regulate mechanical and biochemical coupling within cell clusters; both are critical for the coordination of cell collectives and eventual reduction in the appearance of leader cells.
|
It may be hard to believe, but as a film critic, it takes a lot to displease me. If I choose to see a movie, I usually give it the benefit of the doubt, so it takes a real stinker to make it to the bottom of my list. These are 10 worst movies I had the misfortune of seeing in 2015.
You'll find this movie on some critics' best lists because of its simple and, for them, effective execution of horror movie tropes. But whether you find it terrifying or not likely depends on where you saw it.
In the theater, with an audience primed for scares, you're more likely to go along with the idea of a slow-moving villain being horrific. At home, it's easier to see the film's flaws, and that the threats it contains are actually pretty laughable.
SNL alum Kristen Wiig is attempting to prove she's more than just a comedian, throwing in films like this and The Skeleton Twins that let her show off her dramatic range.
Although she delivers a decent performance, Welcome to Me is a disaster from beginning to end. A lack of character development prevents you from caring about anybody in the entire film, which is a shame since it contains a slew of recognizable actors.
Here, a solid idea of having a stoner as a sleeper CIA agent gets blunted by poor storytelling. Jesse Eisenberg and Kristen Stewart are fine actors in their own right, but even they can't save a film that seems to favor gore over plot.
If the filmmakers had just kept the movie funny with the occasional violent outburst, it would have been fine. But when the laughs get stopped dead in their tracks by having someone's head get blown off — literally — you have to wonder what they were thinking.
The recent wave of young adult books set in dystopian worlds and their subsequent movie adaptations is one that can't be sustained, as the sequel for The Maze Runner and the final Hunger Games showed this year.
But the worst of the bunch was Insurgent, the sequel to Divergent, which played more as a parody of the genre than an exemplar of it. Overly complicated plot machinations, cheesy special effects, and an uninteresting story make for a hopeless film.
There was a time in the not-too-distant past when director Michael Mann could be counted on to deliver a great movie every time. As Blackhat proves, that time has passed, and it may never return.
Mann and his team attempt to goose up a story about a hacker by turning it into an action thriller, a choice that leads to a waste of good actors and some of the most nonsensical action scenes this side of an Expendables film.
Some of the best parts of the Despicable Me franchise are when the Minions are put front and center. But as with most supporting characters, a little of them goes a long way, and devoting a whole movie to them is just way too much.
With the characters' weird noises and oddly multilingual utterances quickly going from cute to repetitive, and a story that doesn't seem to know how to have any fun, the film never gets going. With other great animated options out there, why would anyone waste their time with this drivel?
4. Jupiter Ascending
The Wachowski siblings made their name with The Matrix, which still deserves its classic status. Unfortunately, everything they've done since has fallen into the category of either "meh" or "what is this crap?"
Jupiter Ascending lies squarely in the latter category, as it contains spectacular visual imagery backing up a horrendous story, and acting so bad that it almost overshadowed Eddie Redmayne's Oscar-winning performance in The Theory of Everything.
What is it about the Fantastic Four that makes it impossible to construct a good movie? There have been three made in the last 10 years, and not one of them has turned into something watchable.
This one skips the part about superheroes learning to use their newfound powers in favor of a bunch of unnecessary personal drama that does nothing to advance the story. With a big bad villain who's neither big nor bad, and a quartet of heroes who are nearly impossible to root for, this adaptation is DOA.
This is a movie that most of you have likely never heard of, much less seen, and that's for good reason: It's god-awful. Starring Samuel L. Jackson as the President of the United States, it contains nearly every bad action movie cliche imaginable.
It has it all: horrible one-liners, impossible physical maneuvers, poor special effects, and an ending that goes so far beyond expectations that even Michael Bay would look at it and say, “You might want to tone it down a bit.” The only thing more shocking than the awfulness of the movie is the level of talent that agreed to participate in it.
Oh, Cameron Crowe, where have you gone? The writer/director known for classics like Say Anything..., Jerry Maguire, and Almost Famous hasn't made anything close to a good movie in 15 years, with Aloha representing his nadir.
Another waste of good actors, the film never takes the time to explain anything of import to the story. Crowe hop-scotches between subplots seemingly at will, and one bad performance by Emma Stone — not to mention an ethnically questionable one — overrides any positive attributes. Aloha means "say goodbye to the Cameron Crowe you knew and loved."
|
LSU lineman leaving school for the NFL
Friday
Jan 11, 2013 at 8:55 AM
Les EastThe Advocate
BATON ROUGE — LSU’s list of underclassmen entering the NFL draft reached 11 Wednesday when offensive tackle Chris Faulk decided to forgo his senior season. Faulk was considered a candidate for early entry before the season, but he suffered a season-ending knee injury in the opener against North Texas. Apparently he felt that he had sufficiently recovered from knee surgery to be able to convince NFL scouts that he’s ready for the next level. LSU sports information director Michael Bonnette said Wednesday night that Faulk told him he was going to the NFL. Faulk’s injury started a series of changes on the Tigers offensive line that contributed to an uneven offensive performance as LSU finished 10-3.Later guard Josh Williford was sidelined by a concussion and tackle Alex Hurst left the team for personal reasons. Senior Josh Dworaczyk moved from guard to replace Faulk, and Chris Davenport, who was next in line at left tackle, said Tuesday that he was planning to enroll at Tulane for his final season after having graduated from LSU. That leaves the left tackle spot open for competition next season. Next on the depth chart would be freshman Jerald Hawkins, though starting left guard La’el Collins was a highly regarded left tackle coming out of Redemptorist High School in 2011 and could move back there. An incoming recruit or two could also be in the mix. Faulk, an all-state player from Northshore High School in Slidell, redshirted upon his arrival in 2009 and played in 11 games, starting two, at right tackle as a redshirt freshman. In the spring of 2011, Faulk moved to left tackle and earned a starting position. He was named second-team all-Southeastern Conference and was a preseason second-team pick this season. The other underclassmen from this year’s team to declare early are defensive ends Barkevious Mingo and Sam Montgomery, defensive tackle Bennie Logan, linebacker Kevin Minter, safety Eric Reid, cornerback Tharold Simon, running backs Spencer Ware and Michael Ford, and punter Brad Wing. Cornerback Tyrann Mathieu, who was kicked off the team during preseason camp, is also headed to the draft. It appears all of the Tigers’ potential early entrants, including safety Craig Loston who plans to come back to school, have announced their intentions. The deadline for applying is Jan. 15, but NFL spokesman Greg Aiello said players who apply have 72 hours to rescind their declaration after the filing deadline. He said the league likely will announce the final list of early entrants Jan. 19. Though the 11 early departures are believed to be the most from one school in one draft, ESPN college football studio host Rece Davis said Wednesday that the distinction should be viewed as a positive for the program. “There is great talent in the program at LSU,” Davis said during a telephone interview to promote the Capital One Cup competition. “In addition to aspiring to win championships for their university and their teammates, (top players) aspire to achieve the highest personal goals. “I don’t think it’s bad. Excellence begets excellence. I think it signifies that LSU is good at developing players and getting them ready for the NFL. There’s some great talent whose time to move on has come. It’s a good problem to have. The last thing you want is to have a bunch of seniors coming back who aren’t very good players.” The Capital One Cup standings through the fall season were released Wednesday and will be updated at the end of the winter and spring seasons. LSU is not in the top 10 among the men or the women. Alabama, Indiana and North Dakota State are tied atop the men’s standings and North Carolina and Penn State are tied atop the women’s standings.
|
April 2014
NEW YORK, March 27, 2014 /PRNewswire/ -- XO Group Inc. (NYSE: XOXO), a global media and technology leader and creator of the top wedding website TheKnot.com, today released the results of its annual Real Weddings Study. This comprehensive report surveyed nearly 13,000 US brides and grooms married in 2013. TheKnot.com 2013 Real Weddings Study captures detailed information on the average cost of a wedding, average number of wedding guests, wedding style trends and other key statistics related to weddings in America.
"Wedding budgets have been increasing steadily since the economic downturn of 2008, and in 2013, couples spent a record high average of nearly $30,000," said Carley Roney, cofounder of The Knot. "Couples are more focused than ever on creating a unique, personalized and once-in-a-lifetime experience for their guests – plus they're doing so in a modern way, by planning from their smartphones, publicizing details on social media and more."
GROWING BUDGETS. Couples are spending more on their weddings than ever. Following the trend of growing budgets for the past two years, wedding budgets are at an all-time high with the average wedding cost hitting $29,858, the highest level ever. Less couples (20%) say the economy affected their wedding budget – a statistic that has continually decreased year over year, since reaching an all-time high of 34% in 2009. Wedding standards also continue to rise, in fact, about 1 in 8 couples (14%) spent more than $40,000 on their nuptials, and nearly 1 in 4 (24%) didn't even have a budget.
GUESTS CONQUER ALL.Couples are concentrating on creating an experience for guests. Not only did spend per guest increase to $220, up nearly $26 per guest from 2009 when couples spent $194 per guest, but guest entertainment continues to gain popularity. This rapidly rising category only continues to grow – in 2013, nearly 1 in 3 couples (30%) provided additional guest entertainment, compared with just 11% in 2009. Additionally, while the number of brides who throw a rehearsal dinner, after-party or morning-after brunch has remained steady, brides are spending more money on these events. In 2013, brides spent an average of $1,184 on their rehearsal dinners (up from $1,135 in 2012) and $457 on their morning-after brunch (up from $429 in 2012).
TRADITIONS CONTINUE TO EVOLVE. Couples are foregoing some traditions to modernize their nuptials. Couples are becoming more liberal in their prewedding lifestyles; nearly 3 in 4 couples (74%) now opt to live together prior to getting married. Casual weddings are also on the rise, with 17% of brides classifying their weddings as casual in 2013, up from 12% in 2008. Along the same lines, couples aren't marrying in religious institutions as much as they used to; just 33% chose to do so in 2013, down from 41% in 2009. While the majority of brides still opt to take their spouse's last name (80%), this tradition is becoming less popular, down from 2009 when a whopping 86% changed their last names. Similarly, 1 in 10 (10%) brides kept their names in 2013, up from just 7% of brides who chose to do so in 2009.
SEASONAL SHIFTS. Summer is slowly making its way out of style as the wedding season. Summer is generally defined as wedding season, however, the trend toward beautiful summer weddings is shifting. In 2013, 35% of couples opted for sunshine-filled nuptials, down from 2009 when 41% of couples opted for the season. In recent years, fall has been making headway as the "new" wedding season, but the rise of fall weddings is pausing. In 2013, 32% of couples chose the season, down from its high of 36% in 2011. Spring is gaining traction with couples who aren't afraid of April showers; 26% of brides chose pretty spring weddings, up from 22% in 2011. Winter is also coming from behind to give couples gorgeous snow-filled portraits; though still the minority, 7% of couples chose a winter wedding in 2013, up from 5% in 2012.
PERSONALIZATION AT A PEAK. Weddings are the ultimate form of self-expression. The trend in weddings for many years has been to create a personalized experience, where the couples' personalities are evident in many areas of the wedding. This trend is on the rise, with couples opting for unique details like signature wedding cocktails, which 20% of couples had, up from just 12% in 2008, and hosting interactive food stations (15% of weddings in 2013, up from 13% in 2010). Couples are also personalizing their ceremonies into unique, touching experiences with a family member or friend officiating – just 57% of couples hired a professional ceremony officiant, down from 63% of couples who did so in 2009.
Top 25 Most Expensive Places to Get Married
Based on average cost of a wedding in 2013, couples spend the most on their weddings in the following areas in US.
Figures based on respondents who hired a professional vendor for the service.
Notable Regional Differences
TheKnot.com 2013 Real Weddings Survey can be broken down on a region-by-region basis across more than 80 markets. Regional highlights include:
Cost of Weddings:New York City (Manhattan) and Long Island have the highest average wedding spend ($86,916 and$57,343, respectively), and Utah and Idaho brides had the lowest average wedding spend ($16,816 and 16,159 respectively), excluding honeymoon.
Marriage Age: The NYC Outer Boroughs and Nevada have the oldest brides (33.1 and 32.3 years respectively), whereasUtah and West Texas have the youngest brides (25.7 and 26.7 years), on average.
Engagement Length:Idaho and Rhode Island brides have the longest engagements (21.8 and 20.3 months respectively), while brides in Wyoming and Utah have the shortest (10.5 and 10.6 months), on average.
Wedding Size: On average, Nebraska and Iowa brides have the largest number of wedding guests (222 and 195 respectively), and Nevada and Hawaii brides still have the smallest number of wedding guests (63 and 81).
Wedding Style:Savannah, GA, and Hawaii have the most casual weddings (41% and 38% respectively), while Manhattanand North/Central NJ had the most formal (black-tie) weddings (40% and 35% respectively).
Wedding Dress Spend: Brides in Manhattan and Long Island, NY, spend the most on their dresses ($3,027 and $2,160respectively), while Alaskan and Oklahoman brides spend the least ($804 and $859 respectively).
Honeymoon: Couples from West Virginia (89%), Utah (88%) and Tennessee (88%) are most likely to book a honeymoon, while couples in Alaska and Nevada are still least likely to honeymoon (43% and 44% respectively).
Gift Registry: Brides in Iowa and Nebraska are the most likely to register for wedding gifts (98% and 97% respectively), while brides in Hawaii and Nevada are least likely to register (62% and 69% respectively).
About TheKnot.com 2013 Real Weddings Survey The 2013 Real Weddings Survey captured responses from nearly 13,000 US brides married between January 1 and December 31, 2013. Respondents were recruited throughout the year from TheKnot.com and WeddingChannel.commembership, which represents nearly 80% of brides nationwide. Brides received survey invitations shortly after their wedding date and were asked a comprehensive series of questions about the event and their planning process. All qualified respondents attested that they:
Were 18 or older
Had a wedding in 2013
Survey respondents represent a variety of ethnicities and educational and income levels, and are geographically dispersed across the country. The survey and data management partner for the study was Decipher Inc.
|
News
Hurricanes News
Canes Reduce Training Camp by 10 Players
by
Ken Preston
/ Carolina Hurricanes
RALEIGH, NC – Ron Francis, Executive Vice President and General Manager of the National Hockey League’s Carolina Hurricanes, today announced that the team has trimmed its training camp roster by 10 players. It also was announced that center Jordan Staal has been placed on injured reserve. With the moves, just 25 players remain in the team’s training camp heading into its final two exhibition games, Friday at PNC Arena against Buffalo and Sunday at Washington.
|
Q:
"That" vs "Those" which to use after both a plural and a singular noun?
"The rise of computer technologies and networking is due to collective action similar to that of other social movements, such as the environmental movement, the anti-tobacco movement or the women's movement, for example."
In terms of traditional grammar, is it technically wrong to say "those" instead of "that"? 2. If so, is it still wrong even in colloquial situations?
Are there other errors or slight incoherency from the viewpoint of a native speaker?
FYI since I'm not a native, I would very much like a response from a fluent native speaker.
A:
For your first question, it is technically wrong to say those. However, many people will prefer those to that because it seems to agree with 'other social movements'. But in your sentence, the word that agrees with collective action as it should.
In colloquial situations, either will probably go by unnoticed.
For your second question, the combination of such as and for example is redundant. If I were you, I would leave the such as in and take out for example. I'd also stick in an oxford comma after anti-tobacco movement, but that's a matter of preference.
|
Monday, 22 February 2016
Seconds
This is a film I’ve been wanting to see for years now and I kept meaning to pick up the US Criterion Edition for a while. However, there are a couple of DVD/Blu Ray labels here in the UK that have come into their own in recent years as a credible and, only very slightly, less expensive alternative to Criterion editions and I thought it would be better getting this new edition from The Masters Of Cinema, who seem to have improved considerably over the space of the last ten years in their excellence at acquiring or restoring decent prints and transfers of the titles they put out.
Now I’m not the biggest fan of Rock Hudson and nor am I much familiar with his work. I must have seen a couple of his films with Doris Day when I was a kid and my mother used to avidly watch him on TV in MacMillan And Wife during the 1970s. However, to me he will always be linked in my mind to the production I saw him in as a 12 year old, the 1980 TV mini-series adaptation of Ray Bradbury’s The Martian Chronicles. So even though he’s not exactly done much science fiction like some of his contemporaries were doing around the same era (such as Charlton Heston and Yul Brinner), it’s primarily in this kind of speculative story that I know him as a performer.
Now Seconds is a film which, I have to say I find a bit hit and miss during its almost two hours of running time. When it misses, though, it’s still vaguely watchable and when it hits... well it's really astonishing. I have to say that, for me, the opening of the picture and the first 40 minutes or so of set up, in which Rock Hudson isn’t seen at all, is quite honestly an unbelievably cool and manipulative slice of cinema. This is not to say that there’s anything wrong with Hudson’s performance throughout the film... in fact it’s incredible. However, when his character, or I should say his version of the character, comes into the picture, I did find some of the sequences a bit dull. Still intense but dull... if that’s not too much of a contradiction.
So the film won me over right from the start with a combination of two powerful forces... Saul Bass and Jerry Goldsmith.
Bass’ opening title design is absolutely spooky, as parts of the human face are distorted and stretched in a similar fashion to the way one might perceive one’s own image in a crazy, fun house mirror. It’s all very big, surreal and in your face and I can only imagine how powerful it must have been sitting in front of a huge cinema screen (as they mostly were in those days... huge compared to the majority of those we have now) and being confronted with this. Probably both amazing and seriously disconcerting in equal measure. As the credits progress, different sections of distorted face or head are pitched together in split screens, thus confronting any viewers with an even more unsettling attack on the collective windows of the audience’s soul.
Hand in hand with this is Jerry Goldsmith doing a big sinister piece which starts off gently with violins and small statements, like a chamber concerto, using phrases and orchestration which I personally would associate with generally evil music.... although I’m not nearly enough of a technical expert when it comes to music to know whether the composer used the ‘devil’s interval’ in this sequence. However, to add the complete sense of sinister doom, he then starts bringing in organ tones which certainly did the trick with my dog, who decided it was time to leave the living room at this point. All I know is Goldsmith’s score is a powerhouse along with the Saul Bass titles and, although the film is sparsely but effectively spotted (a Goldsmith trait when the studio allowed for that), it’s an amazing and less than restful score that really heightens the sense of paranoia and tension which I believe the director was probably going for.
Goldsmith’s score continues in the opening scenes of a man being followed by another in a busy train station... although since it’s obviously the beginning of the line, it’s surprising that the passengers chose to sit next to each other in rush hour and were kind enough to leave the main protagonist, as played by John Randolph at this point in the film, a double window seat. However, the sequence astounded me and I was getting very excited because the director was using all kinds of different angles on various moving shots edited rapidly together including... that thing they do where they strap a camera to either the front or back of an actor (both in this case) and film so that the actor’s head or body is kept centre and everything else appears to be moving around him as he walks. Seriously? This seemed out of place and amazing because I thought this was a relatively new approach to shooting, within the last 20 years or so... but here you have John Frankenheimer using it in 1966. So I’m wondering just when this technique started to be used... presumably when industry standard cameras were actually small enough to be able to do this with... and in which film it was first used. Any readers who know the answer to this, please let me know via the comments section for this post.
The film itself concerns a very rich section of the population who want to fake their own death in an elabourate way using a body and major plastic surgery and organ modification (including the vocal chords etc) and our main protagonist is one such person. However, he is drugged and in a wonderfully surreally staged ‘dream sequence’ he is forced to ‘pseudo-rape’ a woman (a company employee) and filmed as insurance that he won’t back out of the deal at some point. The black market but highly successful company which do all this are very thorough in their methods and there’s another wonderfully inventive moment when we see Randolph’s character sign on the dotted line and when we cut to the pen signing it, it’s actually the surgeon’s hand marking up his flesh on the operating table. This is a truly great film for inventive camera work and editing. Although it’s sometimes obtrusive to the illusion of immersion, I found the film to be worth sitting through some of the slightly duller moments for, than I would normally be.
There’s another thing Frankenheimer does early on which is use static cuts of the same character but from longer distances each time, rather than just zoom out. My guess is he does it because he wants to capture the sheer anxiety of the character and slam it home quickly, rather than allow the audience to pull back from the intensity slowly. Whatever the reason he does this though, it’s a pretty fantastic and effective way of conveying the emotions of the actor without diluting it with camera movement. The edit slams you around instead.
When Rock Hudson’s ‘reborn’ version of the character comes into the story as one of the ‘Seconds’ of the title, it begins to come to light that maybe he didn’t make the right decision after all and he’s either a grumpy, unsmiling character or he gets into situations where he goes totally over the top in the other direction. Hudson’s performance is, as I said, flawless but the two party scenes, one in his new home and another outside where he and many others go naked and get in a big tub to crush wine grapes, are actually a little over long and draggy, as far as I’m concerned. And, yeah, I really didn’t expect myself to be saying a sequence which has so many lovely ladies shedding their clothing and inhibitions is a little on the dull side but... I guess every person’s viewing experience is different and that was mine.
Now, there’s an end twist and I thought I knew what it was after about halfway through the movie due to a sequence where Hudson is greeted by an ‘old friend’ at the airport. However, the actual twist, when it comes at the end of the movie, is a lot more pedestrian than I’d hoped for but it doesn’t completely let the movie down and it’s not a bad ending... just a little less than I’d expected from the powerful and claustrophobic set up. Also, I’m not sure but I think the final moments of the movie reveal a flaw in that the person who greets Hudson at the airport in the earlier sequence of the movie now seems to be a bit more of a puzzle than I’d originally thought... but I don’t want to go into why too much, on here, because I want to avoid going into spoiler territory with this one.
So that’s that. At the end of the day, I think Seconds is a movie that genre fans might be interested in. It’s a bit of a flawed masterpiece, it has to be said, but it’s got some really interested visual devices used throughout, has a title sequence by the grand master of title sequence design (not to mention graphic design in general) and a low key but ultimately threatening and powerful score by one of the greatest composers of his generation. So film enthusiasts really should take a look at this one if they haven’t already. Inventive, bleak and with naturalistic performances... a nice little movie, for sure.
|
Blessing for IBC Breakfast March 14 2015
Father in heaven, we are grateful for being able to gather here this morning to share our lives with each other in friendship, love and hope. Please bless those classmates who are unable to attend and heal those classmates who are ill.
We look forward to the spring where your flowers will be dressed in their splendid finery as only you can create them.
Please bless the food as we break bread together, and bless those who have prepared and served it.
Finally we hold up to you, Father, the Hummel family with their loss of Ann. She bravely fought a good fight and is now in your hands.
|
//
// Prefix header for all source files of the 'ProductInfoUtil' target in the 'ProductInfoUtil' project
//
#ifdef __OBJC__
#import <Foundation/Foundation.h>
#import <UIKit/UIKit.h>
#endif
|
Q:
Javascript include namespace
I've a file foo.js with a namespace like this:
// foo.js
const Foo = {
hello: () => {
console.log('Hello world!');
}
}
module.exports = Foo
Now in my file imports.js i can do this:
// imports.js
global.Foo = require('./foo.js');
so in any js files I can call Foo.hello();
Now the question is: in a specific js file is there a way to call hello() without Foo, like using namespace Foo of C++?
A:
Just use require('./foo.js').hello like this:
// In Imports.js
global.Hello = require('./foo.js').hello;
Or if you need to use other things in foo, just create an alias
global.Foo = require('./foo.js');
global.Hello = Foo.hello
Or if you want to use imports, import hello directly:
import {hello} from './foo.js'
Importing it can have space advantages if you're using tree shaking in a build step, because it doesn't have to include unused exports in your bundle.
|
/*****************************************************************************
* ExploringSfMWithOpenCV
******************************************************************************
* by Roy Shilkrot, 5th Dec 2012
* http://www.morethantechnical.com/
******************************************************************************
* Ch4 of the book "Mastering OpenCV with Practical Computer Vision Projects"
* Copyright Packt Publishing 2012.
* http://www.packtpub.com/cool-projects-with-opencv/book
*****************************************************************************/
#include <stdio.h>
#include <iostream>
#include "opencv2/core/core.hpp"
#include "opencv2/features2d/features2d.hpp"
#include "opencv2/highgui/highgui.hpp"
#include "opencv2/imgproc/imgproc.hpp"
#include "opencv2/calib3d/calib3d.hpp"
#include "opencv2/video/tracking.hpp"
#include <fstream>
using namespace std;
using namespace cv;
#include "Distance.h"
|
The Buck Stops Here
Saturday, January 28, 2012
Don't Believe the "Defenders" of Teachers: Teachers Do Matter
You often see education commentators trying to suggest that bad school performance is almost entirely the fault of poverty and other external factors, not the fault of poor teaching. In making this claim, commentators often point to variance in student test scores that is allegedly "explained" by teachers. For example, Anthony Cody says, "Even Eric Hanushek, the economist who has done more to advance these evaluation systems than anyone, admits that teachers only account for around ten percent of the variability in student test scores."
Family and income are surely important, but the "10% of variance" argument is wrong for at least two reasons:
First, in statistical terms, saying that teachers account for 10% of the variance in student test scores does NOT mean that teachers are unimportant. Wrong, wrong, wrong. (At the end of the blog post, I say more about what explaining variance means.)
The eminent Harvard professors Rosenthal and Rubin explained this in a 1982 article, "A Simple, General Purpose Display of Magnitude of Experimental Effect," Journal of Educational Psychology 74 no. 2: 166-69 (that article isn't available online, but is described here).
As luck would have it, Rosenthal and Rubin address the precise example of a case wherein 10% of the variance was explained:
We found experienced behavioral researchers and experienced statisticians quite
surprised when we showed them that the Pearson r of .32 associated with a coefficient of determination (r2) of only .10 was the correlational equivalent of increasing a success rate from 34% to 66% by means of an experimental treatment procedure; for example, these values could mean that a
death rate under the control condition is 66% but is only 34% under the experimental condition. We believe . . . that there may be a widespread tendency to underestimate the importance of the effects of behavioral (and biomedical) interventions . . . simply because they are often associated with what are
thought to be low values of r2.
By analogy, saying that teacher quality explains 10% of the variance would be equivalent to saying that teachers can raise the passing rate from 34% to 66%. That's nothing to sneeze at, and it certainly isn't a reason for teachers to throw up their hands in dismay at the hopelessness of their task.
Second, the fact that teachers account for 10% of variance NOW, given a particular set of data points, tells us little or nothing about the true causal importance of teachers. As Richard Berk explains in his book Regression Analysis: A Constructive Critique, "Contributions to explained variance for different predictors do not represent the causal importance of a variable." 10% isn't a Platonic ceiling on what teachers can accomplish, and the proportion of variance explained tells us very little about how much impact teachers really do have.
A simple hypothetical example makes this clear: Imagine that all teachers in a school were of equal quality. Given equal teachers, any variation in student test scores would automatically have to arise from something other than differing quality of teaching. So a regression equation in that context might tell us that demographics explain a huge amount of the variation in test scores, while teaching quality explains nothing. But it would be completely wrong to conclude that demographics are inherently more important than teaching quality, or even that teaching quality doesn't matter. The exact opposite might be the case, for all that such a regression could tell us.
Moreover, if all teachers became twice as effective as they are now, there would still be variance among teachers and variance among student test scores, and teachers collectively might still "account" for a "small" amount of variance, but student performance might be much higher. The fact that teachers account for 10% of variance today (as large as that actually is) simply does not give us any sort of limit on how much student achievement could rise if the mean teacher effectiveness shifted sharply to the right.
So the would-be defenders of teachers can breathe a sigh of relief: value-added modeling might still be a shaky idea for several other reasons, but there's no need to denigrate the potential of teachers.
What does this mean? The denominator is calculated by taking all the individual Y's (in the education context, all of the student test scores that you're trying to explain), subtracting the average Y value, squaring all of the differences, and adding up all of the squared values. In the context of the following graph, the denominator gives us a measure of the total squared distance (in the vertical direction) that all of the red dots deviate from the average Y value.
The numerator tells us how far the regression line deviates from the predicted Y values. The regression line predicts that the Y values will be along the line itself, which obviously isn't exactly true. So the predicted Y values (that's what the little ^ sign over the Y means) have the average Y value subtracted, the difference is squared, and then all the squared differences are added up.
All in all, the "proportion of variance explained" figure is just a way to represent how close a regression line based on X will come to the actual red dots in the graph, compared to how close a line based on just the average red dot will come.
For the same reason that correlation is not causation, accounting for variance does not provide an upper limit for the true causal importance of a variable. As noted above, the level of variance "explained" is a bad way to determine how important X actually is. See D'Andrade and Hart, for example.
UPDATE: See Cosma Shalizi's post on explaining variance.
Jeff Buckley's "Corpus Christi Carol"
Tuesday, January 17, 2012
The New Groupthink
Our schools have also been transformed by the New Groupthink. Today, elementary school classrooms are commonly arranged in pods of desks, the better to foster group learning. Even subjects like math and creative writing are often taught as committee projects. In one fourth-grade classroom I visited in New York City, students engaged in group work were forbidden to ask a question unless every member of the group had the very same question.
I'm not sure why group seating in classrooms seems to have caught on so strongly. As a parent, I know that children are better behaved (if only by necessity) when they're not sitting close enough to bother someone else, mark on the other child's paper, etc.
Saturday, January 07, 2012
A Case Study in Bias
Two studies came out comparing the performance of schools or teachers.
In the first case, Raj Chetty, John Friedman, and Jonah Rockoff came up with just about the most extensive and sophisticated study of teachers' value-added that I've ever seen. As highlighted in the New York Times, the study includes estimates for how much high-quality teachers improve their students' income years later, and also (see pp. 29 ff.) includes a new way to check for bias by looking at how cohorts of students change performance when a high or low value-added teacher arrives from somewhere else. Very cool.
But such a study, implying that some teachers are better than others, and that teacher quality can be revealed by how well their students do on tests (conditioning on prior achievement and student demographics), is disturbing to some people. Diane Ravitch tweeted at least 67 times the day the study came out, trying to undermine the study by questioning its lack of peer review (so far), the way in which it was conducted, and the very project of looking at test scores in the first place.
In the second case, there's a group called Educate Now in Louisiana that released a PDF chart (available here) that merely lists the schools in New Orleans identified by whether they are Recovery School District schools or voucher-accepting private schools, and then listing what percentage of students score above basic on English and Math in grades 3-5. That's all. No attempt to control for the individual students' prior achievement, no attempt to control for any student demographic variables such as poverty, no attempt to control for the fact that students are eligible for vouchers only if they had been attending a failing public school, no statistical analysis whatsoever.
This is as primitive as it gets, and is a horrible way to judge the merit of voucher schools (as I explained here).
Did Diane Ravitch tweet 67 times criticizing this purported attempt to compare voucher schools to public schools? No: right in the midst of her incessant criticism of an immeasurably superior study, she sent out one tweet that said, "How did voucher schools in New Orleans do?" followed by a link.
Ravitch here displays the worst sort of intellectual bias: when what looks like one of the best studies out there doesn't fit her ideology, she acts as if it is far more questionable than the baloney that she otherwise is happy to plug. To be sure, it's OK to ask questions about the new value-added study, what it means, how it was done, and whether it was oversold in the media. But it's not OK to pass along a worthless analysis of the merits of vouchers.
Anti-reformers need to think a bit more carefully about whether they want someone as their standard-bearer who doesn't know the difference between good and bad research (or, worse, who doesn't care).
|
using Microsoft.WindowsAzure.Storage;
using Microsoft.WindowsAzure.Storage.Table;
using System.Threading.Tasks;
namespace FunctionAppSample
{
public class TableStorage<T> where T : TableEntity, new()
{
private CloudTableClient _tableClient;
private CloudTable _table;
protected TableStorage(string connectionString)
{
var strageAccount = CloudStorageAccount.Parse(connectionString);
_tableClient = strageAccount.CreateCloudTableClient();
}
protected async Task InitializeAsync(string tableName)
{
_table = _tableClient.GetTableReference(tableName);
await _table.CreateIfNotExistsAsync();
}
public static async Task<TableStorage<T>> CreateAsync(string connectionString,string tableName)
{
var result = new TableStorage<T>(connectionString);
await result.InitializeAsync(tableName);
return result;
}
public async Task AddAsync(string partitionKey, string rowKey)
{
if ((await FindAsync(partitionKey, rowKey)) != null) { return; }
var newItem = new T() { PartitionKey = partitionKey, RowKey = rowKey };
await _table.ExecuteAsync(TableOperation.Insert(newItem));
}
public async Task UpdateAsync(T item)
{
await _table.ExecuteAsync(TableOperation.InsertOrReplace(item));
}
public async Task<T> FindAsync(string partitionKey, string rowKey)
{
var ope = TableOperation.Retrieve<T>(partitionKey, rowKey);
var retrieveResult = await _table.ExecuteAsync(ope);
if (retrieveResult.Result == null) { return null; }
return (T)(retrieveResult.Result);
}
public async Task DeleteAsync(string partitionKey, string rowKey)
{
var item = await FindAsync(partitionKey, rowKey);
if (item == null) { return; }
var ope = TableOperation.Delete(item);
await _table.ExecuteAsync(ope);
}
}
}
|
Q:
SCCM Application Deployment - No Status or Errors
Within the Configuration Manager console I have created an application package for WMF5.1 that has a dependency on .NET Framework 4.7.2. It also supersedes WMF3.0 so there's an action included to uninstall that if present through a PowerShell script detection method. The deployment also has an OS requirement for Windows Server 2008 R2 64-bit.
I've deployed this to a device collection which has one Windows 2008 R2 64-bit server in it (Yes the package uses the 64-bit installer). This server has the SCCM agent installed and network connectivity is working fine.
There is no timing restriction on deployment, it's just set to "as soon as possible".
All boundries groups are configured correctly and have deployed to this server and many others in the same subnet before without issue.
Upon checking the monitoring tab for this deployment a day later, The pie chart is completely grey. I've run a summarization many times and all status tabs for this deployment just show the message "Status information is currently unavailable for this deployment"
I've checked a few server logs in the SMS_CCM\Logs folder but not seeing anything obvious. I'm assuming this hasn't reached the client yet seeing as the deployment doesn't look to have detected the device in the collection yet!?
I feel like I'm missing something here but even if my dependency or supersedence rules were wrong, surely the deployment would error rather than just show nothing!
I'm struggling to see how I can troubleshoot something that doesn't appear to have logged any activity for this action.
Any ideas?
A:
I've been looking further into this and even though I had no errors to go by, noticed that the application deployment was set to "Available" rather than "Required" under the "Purpose" drop-down box. Since changing that, the deployment has proceeded and is now installing as expected.
|
A400 Men's Compression Sleeveless Top
Benefits & Features
Designed for training and competing in any sport, SKINS A400 Men's Compression Sleeveless Tops feature dynamic gradient compression. This unique technology accelerates your blood flow to deliver more oxygen to your core muscles – for more power and endurance.
With 400 fit and dynamic gradient compression, you'll notice improved core body control and power, as well as less post-exercise muscle soreness.
Biomechanically placed Memory MX fabric over the spine and around the scapula and obliques provides constant, controlled compression and unrestricted movement.
Sneak a SKINS A400 Men's Compression Sleeveless Top under your team shirt to fire up your body like nothing before.
How do you rate this product?
*Name
*Email (not shown on site)
*Review
Travis
·
5/02/2015
I was initially a bit sceptical when i ordered these for golf, however, i've never been more pleased to be proven wrong! They wick sweat fantastically, as well as helping to stabilise my core at the top of my backswing. Heartily recommended!
Tim
·
16/07/2010
i like the sleeveless top works very well, love the neck area how it is soft and feels comfortable unlike other brands. would prefer to have slightly more compression around the stomach area though.
3D body scan data and 400 key fitting points have been used to create a more precise and comfortable fit to maximise compression benefits.
SKIN fit
SKINS unique sizing system is based on our unique Body Mass Index (BMI)/anthropometrical algorithm, because you only get the full benefits of gradient compression if your SKINS fits you properly
Muscle Focus
Biomechanically positioned panels and seams support and stabilise the muscles to reduce fatigue and risk of injury.
Fabric Technology
Moisture Management
Moisture wicking fabric draws sweat away from the body to the surface of the garment where it evaporates to keep you dry.
Odour
Permanent antimicrobial treatment discourages bacteria by ionically absorbing the odour causing substances normally associated with Lycra wear. These features are built into all our fabrics, so they can’t ever be washed out.
50+ UV Protection
SKINS proven 50+ UV protection.
Memory MX
Memory MX Fabric has a superior high-stretch yarn that doesn’t change shape or sag, so you maintain consistent compression, session after session. A400 Long Tights have 70D Memory MX over the knee and Achilles tendon for unrestricted natural movement, and 210D Memory MX over the ITB and TFL muscle groups to help reduce the risk of injury.
Warp Knit
SKINS compression fabric is made from a warp knitted mix of yarns that give you a carefully calculated level of stretch for controlled compression and durability.
|
golang.org/x/sys v0.0.0-20190912141932-bc967efca4b8 h1:41hwlulw1prEMBxLQSlMSux1zxJf07B3WPsdjJlKZxE=
golang.org/x/sys v0.0.0-20190912141932-bc967efca4b8/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
|
Monday thru Friday at
www.DailyCatholic.org
See why so many consider the Daily CATHOLIC as the"USA Today for CATHOLICS!"
e-mail:[email protected]
WEDNESDAY
July 7, 1999
SECTION ONE vol 10, no. 130
To print out entire text of Today's issue, print this section as well as SECTION TWO
Because of his background and the fate of his parents, Cardinal Jean-Marie Lustiger has long been a champion of human rights worldwide while serving as Archbishop of Paris for 18 years
Today we feature our seventy-fourth red hat in alphabetical order. He is the 73 year-old prelate from Paris Cardinal Jean-Marie Lustiger, who was born of Polish Jewish parents in Paris but preserved from the Holocaust when his folks were extradited and eventually executed at Auschwitz in 1943. He converted to Catholicism and became a priest always upholding the cause of human dignity. For more on this Jewish-born Prince of the Church who was elevated to the cardinalate in Pope John Paul II's Consistory of February 2, 1981, click on COLLEGE OF CARDINALS COLLECTION
74. Cardinal Jean-Marie Lustiger
Like recently canonized holocaust victim Saint Edith Stein, the Archbishop of Paris Cardinal Jean-Marie Lustiger is a bridge between the Jewish peoples and Christians for he was born of Jewish parents in Paris, France on September 17, 1926. His parents, of Polish descent, were killed in the Nazi death camp of Auschwitz in 1943 during World War II. But when they were deported from Paris early in 1940, young Jean-Marie was mercifully spared and taken in by a Catholic family in Orleans. That traumatic event changed his life forever. So taken was he by the faith of his new foster parents that he converted and was baptized into the Roman Catholic Church on August 25, 1940 at the age of 14. As a young college student after the war he became very active in Young Christian Students in Paris and it was during this time he realized a vocation to the priesthood. After studies at the Sorbonne where he obtained his degrees in the humanities and philosophy and the necessary major scholasticate studies, he was ordained a priest on April 17, 1954.
After some more studies and pastoral work he was appointed the Cure of St. Jeanne de Chantal in Paris where he dedicated his time night and day to guiding his parishioners with a special emphasis on the youth and the elderly, always striving to honor the human dignity of every person he came in contact with. He became legendary with church-goers everywhere in France for his dynamic preaching. It also attracted the attention of Pope John Paul II who appointed him the Bishop of Orleans on November 10, 1979. He was installed a month later on the Solemnity of the Immaculate Conception. Two years later the Holy Father elevated him to succeed Cardinal Marty as Archbishop of Paris on January 31, 1981. Two years later the Pope honored him by naming him in his February 2, 1983 Consistory, bestowing the red-hat on Cardinal Lustiger at the age of 57. He received the titular church of St. Louis of France. Shortly thereafter he was appointed Ordinary for all French Eastern-Rite faithful who did not have ordinaries of their own.
He has produced numerous important documents and significant homilies in France and globally and has come to be respected for his strong stance on human dignity, a refrain campaigned vigorously by another Pole - His Holiness John Paul II. The Holy Father has shown confidence in Cardinal Lustiger by appointing him to membership in various Curias including the Secretariat of State in the second section, the Congregation for Divine Worship and Sacraments, the Congregation for Bishops, the Congregation for Oriental Churches, the Congregation for the Clergy and the Congregation for Institutes of Consecrated Life and Societies of Apostolic Life as well as the Pontifical Council for Culture. Though he, along with fellow French colleagues Cardinals Roger Etchegarry, Pierre Eyt and Paul Poupard are all above 65, they are not expected to be selected as serious candidates in the next papal conclave, but their influence could have a strong swing vote in the process.
Reviewing the first six hundred years of the Church as we set sail on the Barque of Peter from the time of Christ to Saint Gregory the Great
With the summer mode in vogue, we suspend this series until August, allowing the reader to catch up on past articles they may have missed in archives, including the Establishment of the Blessed Sacrament and the early struggles of the Church in this on-going abridged History of the Mass and Holy Mother Church over a 2000 year span called 2000 YEAR VOYAGE ON THE BARQUE OF PETER. For the first fifteen installments, click on BARQUE OF PETER
Taking time to taste Food for Thought appeteasers over the past year
With summer upon us we follow the summer schedule and bring you retro vignettes from the past year of humorous and moving Food for Thought Appeteasers that could bring a chuckle, a tear or both but definitely will make you stop and think. For past Appeteasers, click on Archives for FOOD FOR THOUGHT APPETEASERS
SIMPLY SHEEN:
Sometimes we think too much!
They say a picture is worth a thousand words, but the words of Bishop Fulton J. Sheen have been known to launch a thousand images in one's mind, one of the ways this late luminary did so much to evangelize the faith. Because of the urgency of the times and because few there are today who possess the wisdom, simplicity and insight than the late Archbishop who touched millions, we are bringing you daily gems from his writings. The good bishop makes it so simple that we have dubbed this daily series: "SIMPLY SHEEN".
"There are always sensitive souls who are much too ready to think evil of themselves and to distress themselves with their evil thinking. It is better not to appraise one's own spiritual life and progress, but to leave it to God and to band all attention to advancing in love of God and love or neighbor."
"I do solemnly tell you that there is no violence committed against any one of My children in My Name that is pleasing before My Throne. Those who use My Name to murder a murderer at any abortion clinic is satan's victim and not My servant! I alone Am Life. I alone give Life. I alone take Life!"
Those powerful, forceful words come from Our Lord on January 22, 1995 during the week a nation celebrated the Roe vs. Wade decision. Jesus' topic in Message #586 was the abominable sin of abortion in which He imparted to the Hidden Flower of the Immaculate Heart that any kind of murder is an abomination. He affirms that violence is not the answer, but rather prayer and penance, as Christ calls on all bishops and priests to return the Most Holy Eucharist to Its rightful prominence in our Church both in placement of the Tabernacle and total reverence and respect for the Blessed Sacrament, where He is present Body and Blood, Soul and Divinity. Jesus asks all His consecrated ones to pray for Divine Mercy on a sinful world. His Blessed Mother prefaces her Divine Son's words in Message #585 four days earlier, calling her little ones to persevere in the fight for the Sanctity of Life and to surrender our own wills to the Divine Will so God can work through and with us in mitigating much of the chastisement ahead in the goal of ultimately ridding the world of evil through the Triumph of her Immaculate Heart. For Messages #585 and #586, click on "I SOLEMNLY TELL YOU..."
Messages 585 and 586
Message Five-Hundred-eighty-five, January 18, 1995
(Imparted to the Hidden Flower by the Immaculate Heart of Mary)
Say now to all the world: I am your Mother and I weep for all who suffer; they bear great pains in body, mind and soul but they do not come to God for help. Their pain grows worse because they seek release from their sufferings through human means.
O! When shall you hear my pleas? When shall you turn back to God and allow the sinful world to pass you by? There is no happiness except in God. He alone can heal; He is the Divine Physician, for He heals the soul before the body, knowing the soul is made for Eternal Life with Him.
How many warnings must my Divine Son give to the world before His children believe and live in a holy way? I, your Mother, do solemnly tell you that the earthquakes which ravage randomly shall soon begin a chain reaction that shall cause worldwide sufferings. I tell you this because the forces of evil are strong and daily they push forward their plans to bring the antichrist to reign.
It is only through constant prayer from the heart, united to my Divine Son's Passion, that many sufferings can be mitigated, the days of suffering shortened, and the Justice swift and merciful.
Do not fear, but do not cease, each of you, to be an apostle in these end times. My Immaculate Heart weeps for you, pleads for you, and loves you.
O! Surrender all to God Who must be your All, and you shall know the truth of my words.
I love and bless you! Thank you for responding to my Call!
Message Five-Hundred-eighty-six, January 22, 1995
(Imparted to the Hidden Flower by the Sacred Heart of Jesus)
On this day, I, Jesus Christ, Only-begotten Son of God, Second Person of the Blessed Trinity and Savior of the world, do solemnly tell the world:
Behold, the chalice of the suffering of the innocents is overflowing. Behold, the stench of the slaughter of the innocents is greater than in that moment in Bethlehem when Herod, My adversary slaughtered the Holy Innocents.
Behold, I, Jesus Christ, do solemnly say to all the world: THE EVIL SIN OF ABORTION MUST END NOW! This evil rocks the Heavens. All the hosts of Heaven weep, and My Own Most Sacred Heart sheds drops of Blood for this outrage!
When shall you hear Me, O world gone mad by satan's designs? When will you once again recognize My Sovereignty as King of Kings, Creator of all life!?! When, O world, shall you cease your abominations? When, O when shall you fall to your knees before My True Presence in the Blessed Sacrament and beg of Me to stop this horrific deed?
I do solemnly tell you that there is no violence committed against any one of My children in My Name that is pleasing before My Throne. Those who use My Name to murder a murderer at any abortion clinic is satan's victim and not My servant! I alone Am Life. I alone give Life. I alone take Life!
Therefore, hear Me, O world, especially all bishops and priests:
The sin of abortion will not stop until all reverence is returned to Me in The Most Blessed Sacrament. I want My churches open at every hour! I want My Tabernacle to be returned to Its rightful place where My children might come to Me. I want to be placed in the Monstrance, and I want Adoration of My True Presence to be a full part of every church.
I want every bishop to announce this in his Diocese and to hear no dispute from any priest or lay committee. I want all bishops and priests to faithfully lead My children in Adoration of The Blessed Sacrament, and I want My True Presence in The Most Blessed Sacrament to be taken in solemn procession to every abortion clinic. There I ask My children, priests, religious and lay, to kneel in prayer and to humbly recite the Rosary and Chaplet of Divine Mercy, so that through Divine Mercy abortion ends.
I do solemnly tell My people: Let My Order of Divine Mercy begin and let its Rule be lived fully. It is through the on-going work I gave to My saint Faustina that I, Jesus Christ, desire for My Divine Mercy to be poured out upon your sinful world. It is My Plan that through this new Institute the world shall come to drink from the Cup of My Most Precious Blood, that My Mercy may be visibly seen and you shall be healed in soul, body and mind.
Hear Me, for soon the world shall be given a great sign of My Presence among you. Let all who have ears hear!
No longer shall abortion be tolerated. No more shall I allow man to commit even more evil in My Name, which is Most Holy.
ENOUGH!
Return to Me, for already My angels stand at the ready to bring the full weight of Justice to the world.
Hear My Words and those of My Holy Mother, and live these Words. I Come! Prepare!
DAILY LITURGY
Today is the Fourteenth Wednesday in Ordinary Time with tomorrow being the Fourteenth Thursday in Ordinary Time. For the readings, liturgies and meditations for these two days, click on DAILY LITURGY.
Thursday, July 8, 1999
Green vestments
PRAYER & DEVOTION
July is the month of the MOST PRECIOUS BLOOD OF JESUS and so, on Ordinary days this month we bring you the Prayer of the Most Precious Blood of Our Lord Jesus Christ:
Almighty, everlasting God, Who didst appoint Thine
only-begotten Son Redeemer of the world, and didst deign
to be appeased by the shedding of His Blood; grant, we
beseech Thee, that by our solemn service we may so honor
what He paid as the price our Redemption, and by Its virtue
be so defended here on earth from the evils of this present
life, that we may forevermore enjoy Its fruit in Heaven.
Throught the same Jesus Christ, Thy Son our Lord, Who
liveth and reigneth with Thee in the unity of the Holy Spirit,
God, world without end. Amen.
Events Today in Church History
On this date in 1304, Blessed Pope Benedict XI, the 194th successor of Peter, was poisoned by a dissident group of Italians led by Nogaret whom the Pope denounced for their sacriligeous crimes in Anagni. Prior to this, the Holy Father had settled a dispute that arose between his predecessor Pope Boniface VIII and the equally stubborn King Philip IV of France. For other pertinent events throughout the centuries that are memorable in Church history today, click on MILLENNIUM MILESTONES AND MEMORIES
Historical Events in Church Annals for July 7:
664 A.D.
Death of Saint Ethelburga,nun and abbess who was also known as Saint Aubierge. She was the daughter of Anna, king of the East Angles. She died in Faremoutier, France.
739 A.D.
Death of Saint Willibald, son of the West Saxon king Richard and Saint Boniface's cousin. Willibald was the first Englishman to make a pilgrimage to Rome in 721 as well as Jerusalem in 724 and Constantinople in 730. Though captured by Saracens, he was eventually released and spent ten years at Monte Cassino in Italy. Then Pope Saint Gregory III dispatched him to Germany where Boniface ordained him and where he ruled as bishop for forty five years before his death at Eichstatt.
1304 A.D.
Death of Blessed Pope Benedict XI, 194th successor of Peter. Benedict was poisoned by eating a tainted fig placed in his dish by conspirators. He was born in Treviso and elected on October 27, 1303. During his one year pontificate he settled a difficult dispute with the king of France Philip IV, also known as Philip the Fair who wasn't exactly fair in his legendary disputes with Benedict's predecessor Pope Boniface VIII who considered the king a mortal enemy. Because he was a just and compassionate man, Benedict raised the ire of the unjust and those who felt they could do whatever they wanted to. This made him a a target of those he excommunicated, including Nagoret and his Italian accomplices who had done unspeakable acts in raiding Anagni. Because of the Pope's proper actions, he was bitterly persecuted by this group of sacriligeous dissidents and murdered.
1456 A.D.
The French ecclesiastical court at Poitiers clear Saint Joan of Arc of all suspicion of heresy, allowing her to lead a crusade to retrieve the besieged city of Orleans.
June 25th Medjugorje Monthly Message
Dear children! Today I thank you for living and witnessing my messages with your life. Little children, be strong and pray so that prayer may give you strength and joy. Only in this way will each of you be mine and I will lead you on the way of salvation. Little children, pray and with your life witness my presence here. May each day be a joyful witness for you of God's love. Thank you for having responded to my call. For more on Medjugorje, click on MEDJUGORJE AND MORE
The DAILY WORD
"These twelve Jesus sent forth, having instructed them thus: 'Do not go in the direction of the Gentiles, nor enter the towns of Samaritans; but go rather to the lost sheep of the house of Israel.'"
|
[Unity] [Programmer] [Composer] Young fresh dev looking for collaborators or a team to join
Hi friendly itch.io folk! I'm a computer science major, classically trained musician, and gamer who wants to start making games. I have a little experience with Unity but I'm down to learn any engine. I'm also comfortable in a range of programming languages, including C#, C, Java, and Python. I can write all kinds of music, from baroque classical to jazz to grunge, so we can work with the best genre for the project.
What I'm looking for in a collaborator: someone who can fill out the other aspects of game development (e.g. skills in art, maybe writing, additional programming) and who's passionate about video games as a form of artistic expression. I don't have a project in mind, so we'll be doing some hardcore brainstorming seshes. I'd also be happy to join any teams working on established projects who need additional hands.
Message me via [email protected] and we can talk about our favorite games and how I'm loving Dishonored 2's level design.
Hey, an artist and I are looking for a programmer and a sfx guy for game jamming. We plan to do the 7DRL challenge starting in a few days, and then doing 1 Game A Month and other jams. If you're interested in joining us, what would you rather do? Programming or sound? I would like to have you as a programming partner, and when we find a sound guy you would be able to give your educated thoughts on his work.
|
Ivan Turner is the designer of the 9 Kingdoms game system. He developed the system over several years, starting with a basic game engine that included moving figures over 3X3 woooden blocks. In addition to his work on 9 Kingdoms, Ivan is a science fiction and fantasy author. 9 Kingdoms itself has its own original mythos which will be published as stories for e-readers and in the print version of the rulebook. His other work can be downloaded onto most e-book reading devices from most sites. His current Zombies! series has become a cult favorite among e-readers. He has also received critical acclaim for Forty Leap, a tale of involuntary time travel.[From http://www.ninekingdoms.com/]
|
Getty Images
Buccaneers coach Bruce Arians was happy that the team would be putting on the pads. He wasn’t happy with everyone’s performance in pads, however.
“It was a little raggedy, more raggedy than I thought it was,” Arians said after practice, via JoeBucsFan.com. “The noise level went up and some guys messed their pants. But we’ll grade the film.”
When Arians grades the film, it sounds like plenty of guys will be getting not an A or a C but an S.
“Mental errors, we had some guys today just crap their pants today in pads,” Arians said.
If a football player is going to crap his pants, it’s better to crap his pants in practice than during a game. The challenge will be to learn from the practice-variety pants-crapping and keep it from becoming an in-game poopfest.
|
DR CROKES HAVE announced that manager Pat O’Shea has stepped down from his position as manager having led the Kerry champions over the course of the last three years.
The 52-year-old led Dr Crokes to the All-Ireland Senior Club Championship in 2017 — just their second ever All-Ireland title and the club’s first in 25 years.
The Killarney native also helped Crokes to two Munster Championships, three County Championships as well as County Club, County League and an East Kerry Championships.
A statement on Saturday confirmed his departure: “The Club announces that Pat O’Shea has stepped down as manager of our Senior Football Team. This was Pat’s own decision and he will be taking up other roles within the Club.”
O'Shea celebrates with Colm Cooper after defeating Slaughtneil in the 2017 All-Ireland Senior Club Football Championship final. Source: Tommy Dickson/INPHO
The side will play in Division 1 of the County League tomorrow against Kerins O Rahllys, with the team set to be managed by Edmund O Sullivan and the rest of the existing management team.
Be part
of the team Access exclusive podcasts, interviews and analysis with a monthly or annual membership. Become a Member
The club’s executive will meet during the next week to formally consider what changes, if necessary, will be made in the management team and will be announcing the result as soon as possible following that meeting.
This weekend’s announcement brings the curtain down on O’Shea’s second spell as manager. The former Kerry senior boss previously led Dr Crokes to successive County Finals in 2005 and 2006.
Murray Kinsella and Andy Dunne dissect Ireland’s disappointing Six Nations campaign, and discuss the pros and cons of rugby’s new law proposals in the latest episode of The42 Rugby Weekly:
Source: The42 Rugby Weekly/SoundCloud
Subscribe to our new podcast, The42 Rugby Weekly, here:
|
1. Field of the Invention
The present invention relates to a controller for controlling the frame refresh rate of an active matrix display. The present invention also relates to a display controller including such a frame rate controller and to an active matrix display including such a controller. Such displays may be used in portable equipment where data may be supplied to the display in a variety of formats and where it is desired to minimise display power consumption.
2. Description of the Related Art
FIG. 1 of the accompanying drawings shows a typical active matrix liquid crystal display of known type. The display comprises an active matrix 1 of N rows and M columns of picture elements (pixels). Each pixel comprises a pixel electrode 2 facing a counter electrode (not shown) with a layer of liquid crystal material (not shown) therebetween. The pixel electrode is connected to the drain of a pixel thin film transistor (TFT) 3, whose source is connected to a data line 4, which is common to all of the pixels of a column, and whose gate is connected to scan line 5, which is common to all of the pixels of a row.
The data lines 4 are connected to a data line driver 6, which receives timing, control and data signals from a data controller (not shown) and which supplies analogue voltages for charging the data lines 4. The scan lines 5 are connected to a scan line driver 7 which is controlled by the timing signals and which supplies scan line pulses to the scan lines 5 one at a time in a cyclically repeating sequence.
Image data are transmitted to the data driver on a frame by frame basis. Within each frame, image data are transmitted line by line with each line of data corresponding to the required display states of a horizontal row of pixels of the display. The lines of data are loaded one at a time into the data line driver 6 which charges the data lines 4 to the required voltages. The scan line driver 7 then supplies a scan pulse to the row of pixels to be updated. The pixel transistors 3 of the row receive the scan pulse at their gates and are switched to a conductive state so that the voltages on the data lines 4 charge the pixel electrodes 2 of the line being refreshed. This is repeated row by row until the whole display has been refreshed by a fresh frame of data. This is then repeated for each frame of data.
FIG. 2 of the accompanying drawings illustrates a typical liquid crystal display controller 10 in the form of an integrated circuit which is generally physically separate from the display. The controller 10 comprises a timing generator 11 which receives clock signals (CKS), horizontal synchronisation signals (HS) and vertical synchronisation signals (VS). The timing generator 11 passes these timing signals to the display and generates timing signals for controlling the operation of the display controller 10.
The controller 10 is capable of receiving video data in either luminance and chrominance format (Y, Cr, Cb) or in RGB (red, green, blue) format. A matrix 12 converts the chrominance format data into RGB format data. An on-screen display mixer 13 receives the RGB data either from the matrix 12 or directly from an RGB input and mixes this as desired with on-screen data from an external static random access memory (SRAM) 14 so that any on-screen display data overwrite the video data. The RGB outputs of the mixer 13 are connected to a gamma correction circuit 15, which compensates for the non-linear response of the pixels to voltage and which allows picture adjustments to be made, for example to the colour, brightness and tint of the displayed image.
The RGB outputs of the gamma correction circuit 15 are supplied in parallel digital format to a digital output 16 for use with displays which require digital input video data. For displays which require analogue input data, the outputs of the gamma correction circuit 15 are supplied to a digital/analogue converter (DAC) 17, which converts the red, green and blue image data to corresponding analogue voltage levels. These voltage levels are amplified by an amplifier 18 and supplied to an analogue output 19.
In typical liquid crystal controller integrated circuits, the frequency of the data can be adjusted to the particular requirements of the display. For example, the controller 10 may output data in either SVGA format or XGVA format, which have different data transmission rates for a given frame rate. The frame rate itself is typically fixed to a frequency which is characteristic of the refresh rate required by the liquid crystal material of the display.
In displays which are for use in portable or battery-powered equipment, it is desirable to reduce the power consumption as much as possible so as to prolong battery life and reduce the frequency of replacing batteries. U.S. Pat. No. 5,926,173 discloses a power saving technique for such a display in which, when new image data are sensed as being supplied to the liquid crystal display (LCD), the power supply to the LCD is stopped. U.S. Pat. No. 5,757,365 discloses another power saving technique for display drivers, in which the absence of image data is also sensed. When this is the case, the drivers, which contain a frame memory, operate in a lower power self-refreshing mode.
U.S. Pat. No. 5,712,652 discloses a portable computer having an LCD. This patent specification discloses reducing the refresh rate of a video graphics controller so as to reduce power but does not describe any technique for achieving this.
U.S. Pat. No. 6,054,980 discloses an arrangement for providing frame rate conversion between a computer supplying display data at one frame rate and a display device which cannot operate at such a high frame rate, but in which the supply and display frame rates are not greatly different from each other. This is achieved by the use of a frame buffer in which image data are written at the supply rate and are read at the display rate so that each (N+1)th frame of image data is effectively dumped, where N is an integer greater than zero.
U.S. Pat. No. 5,991,883 discloses a technique for managing power consumption in laptop computers and the like. The display refresh rate is adapted according to the type of images which are to be displayed. A reduced refresh rate is achieved by reducing the processing speed of image data, for example by reducing the pixel clock rate of a video graphics controller.
U.S. Pat. No. 5,446,840 discloses reducing the rate at which video data are supplied so as to take some of the processing burden off the CPU of a computer system running graphical user interfaces. New video data are written to a relatively fast RAM and then refreshing or updating a display device takes place at a relatively slow rate which is just fast enough to avoid undesirable perceptible visual artefacts.
|
94 F.3d 642
NOTICE: Fourth Circuit Local Rule 36(c) states that citation of unpublished dispositions is disfavored except for establishing res judicata, estoppel, or the law of the case and requires service of copies of cited unpublished dispositions of the Fourth Circuit.Fay G. TOWNS, Plaintiff-Appellant,v.Shirley S. CHATER, Commissioner of Social Security,Defendant-Appellee.
No. 95-3145.
United States Court of Appeals, Fourth Circuit.
Submitted July 25, 1996.Decided Aug. 14, 1996.
Mary V. Rehmann, Morgantown, West Virginia, for Appellant. Charlotte Hardnett, Chief Counsel, Region III, David M. Frazier, Assistant Regional Counsel, Office of General Counsel, SOCIAL SECURITY ADMINISTRATION, Philadelphia, Pennsylvania; William Wilmoth, United States Attorney, Helen Campbell Altmeyer, Assistant United States Attorney, Wheeling, West Virginia, for Appellee.
Before LUTTIG and MOTZ, Circuit Judges, and PHILLIPS, Senior Circuit Judge.
Affirmed by unpublished per curiam opinion.
OPINION
PER CURIAM:
1
Fay G. Towns filed a claim with the Social Security Administration in March 1992 for Supplemental Security Income, alleging disability commencing December 16, 1991, as a result of back pain. After denial and reconsideration, Towns requested a hearing before an Administrative Law Judge (ALJ). The ALJ decided that Towns was not disabled under the Social Security Act because Towns had no non-exertional limitations and although he had a musculoskeletal impairment and could not return to past relevant work, he also had the residual functional capacity to perform at least sedentary work. The Appeals Council denied Towns' request for review. The ALJ's decision then became the Secretary's final decision.
2
Towns filed a complaint in the district court challenging the final decision of the Secretary. The parties consented to a disposition by a magistrate judge pursuant to 28 U.S.C.A. § 636(c) (West 1993). The magistrate judge entered a final order affirming the decision of the Secretary. This appeal followed.
3
We review the Secretary's final decision to determine whether it is supported by substantial evidence and whether the correct law was applied. 42 U.S.C.A. § 405(g) (West Supp.1996); Hays v. Sullivan, 907 F.2d 1453, 1456 (4th Cir.1990). Towns claims that substantial evidence does not support the ALJ's finding of ability to perform sedentary work. However, the ALJ gave specific reasons for his determination and we will not disturb it. Hammond v. Heckler, 765 F.2d 424, 426 (4th Cir.1985). The ALJ made a thorough evaluation of the evidence, and we conclude that the Secretary's decision is supported by substantial evidence and was based on the correct legal standards.
4
Towns also alleged that the ALJ failed to develop a full and fair record and therefore the administrative proceedings were fundamentally unfair. Towns waived his right to appear at the hearing before the ALJ, but only after the ALJ gave him several warnings and suggestions to obtain free legal advice. The ALJ obtained evidence from Towns' treating physicians and from the claims representative who observed Towns and noted his daily activities. In addition, there is no evidence in the record that Towns was prejudiced by a mental impairment.*
5
Accordingly, we affirm the district court's judgment. We dispense with oral argument because the facts and legal contentions are adequately presented in the materials before the court and argument would not aid the decisional process.
AFFIRMED
*
Towns' brief discusses and cites to a psychological evaluation attached to the complaint filed in the district court. This evaluation was not before the Secretary when she made her final decision. Towns' brief also refers to June 1994 back surgery occurring after the Secretary's denial of benefits. This court is restricted only to evidence presented in the administrative record. See Wilkins v. Secretary, Dep't of Health & Human Servs., 953 F.2d 93, 96 (4th Cir.1991)
|
Nelson v New York City Tr. Auth. (2020 NY Slip Op 00671)
Nelson v New York City Tr. Auth.
2020 NY Slip Op 00671
Decided on January 30, 2020
Appellate Division, First Department
Published by New York State Law Reporting Bureau pursuant to Judiciary Law § 431.
This opinion is uncorrected and subject to revision before publication in the Official Reports.
Decided on January 30, 2020
Richter, J.P., Gische, Mazzarelli, Gesmer, JJ.
10914 160417/13
[*1] Ke'Andrea Nelson, Plaintiff-Respondent,
vNew York City Transit Authority, et al., Defendants-Appellants.
Steve S. Efron, New York, for appellants.
Schwartz Perry & Heller LLP, New York (Brian Heller of counsel), for respondent.
Order, Supreme Court, New York County (Kathryn Freed, J.), entered March 22, 2019, which, to the extent appealed from, denied in part defendants' motion for partial summary judgment, unanimously affirmed, without costs.
Defendants New York City Transit Authority (NYCTA) and Metropolitan Transit Authority Bus Company (MTA) failed to establish prima facie entitlement to partial summary judgment dismissing plaintiff's remaining claims of discrimination and retaliation. The record, including the findings of the neutral arbitrator, does not conclusively resolve factual issues regarding MTA's motives in levying various disciplinary charges against plaintiff between 2012 and 2016 (compare Novak v St. Luke's-Roosevelt Hosp. Ctr., Inc., 136 AD3d 435, 436 [1st Dept 2016]; Collins v New York City Tr. Auth., 305 F3d 113, 119 [2d Cir 2002] [decision of a neutral arbitrator upholding charges of misconduct was "highly probative of the absence of discriminatory intent"]). Unlike Collins and Novak, plaintiff here challenged a number of disciplinary charges, including two dismissals, and the arbitrator twice found that defendants' penalties were disproportionate to the charged misconduct.
Further, plaintiff independently lodged formal complaints that her superiors at MTA made discriminatory, disparaging comments about her race and gender before any disciplinary actions were taken against her. Viewed in context, the alleged pervasive, racist, and sexist comments and conduct preclude summary dismissal of this case as "insubstantial" (Hernandez v Kaisman, 103 AD3d 106, 115 [1st Dept 2012]).
THIS CONSTITUTES THE DECISION AND ORDER
OF THE SUPREME COURT, APPELLATE DIVISION, FIRST DEPARTMENT.
ENTERED: JANUARY 30, 2020
CLERK
|
Fintech startup Bnext has raised a $25 million funding round. The Spanish company is building a banking product and has managed to attract 300,000 active users.
DN Capital, Redalpine and Speedinvest are leading today’s funding round. Existing investors Founders Future and Cometa are also participating. Other investors include Enern, USM and Conexo.
When you open a Bnext account, you get a card and you can upload money to your account. Bnext accounts aren’t technically bank accounts — the company has an e-money license. You can then use your card and spend money anywhere around the world without any foreign transaction fee. You also can freeze and unfreeze your card from the app.
“As of now we’ll stick to the e-money license, as our international expansion plans complicate potential passporting of banking licenses. We will first need to understand in which countries makes more sense to get a banking license, and then we’ll make a decision,” co-founder and CEO Guillermo Vicandi told me.
You also can connect to your traditional bank accounts from the Bnext app. This way, you can manage your money from a single app.
And Bnext takes this one step further by offering financial products from third-party companies as well. It’s clear that the company wants to build a financial hub, the only finance app that you need.
You can lend money to small and medium businesses and earn interest through October, you can save money using Raisin, you can get a loan, a mortgage, an insurance product, etc. Bnext generates revenue from those partnerships.
While Bnext only operates in Spain for now, the company has managed to attract 300,000 active users. It processes €100 million in transactions every month ($109 million).
Up next, Bnext plans to offer premium plans with more features and individual IBANs. The company also plans to expand to Latin America, starting with Mexico later this year.
|
###### Strengths and limitations of this study
- Key study strengthsare that we are able to compare overweight trends over time among all age groups, to explore the consumption of ultra-processed foods and physical activity and to investigate hypothesised risk factors for overweight, using regression-based methods.
- We use a recent dataset that is nationally representative of Indonesia.
- A key limitation is that we largely employ cross-sectional data, and therefore, we cannot infer any causal associations.
- Data on the consumption of a select number of ultra-processed foods were only collected in 2014, which limits our ability to make conclusions about the association between ultra-processed foods and overweight over time.
- Physical activity levels were based on respondent recall.
Introduction {#s1}
============
Indonesia is undergoing a nutrition transition as one-third of adults are now overweight or obese. Nutrition transition theory suggests that economic development, urbanisation and globalisation result in an increase in the consumption of ultra-processed foods and a decrease in physical activity,[@R1] which subsequently lead to a higher prevalence of overweight and non-communicable diseases. Context-specific policy, food systems, sociocultural norms and socioeconomics are also thought to play a role.[@R2] Mitigating the obesity pandemic through appropriate programmes and policies requires a better understanding of setting-specific trends and their underlying determinants. Yet, prior studies provide limited information on the changing prevalence and risk factors for overweight, by age group, in Indonesia. We aimed to fill this gap in the literature by detailing overweight prevalence over time and investigating the relation between key risk factors and overweight, among all age groups in Indonesia.
Several factors have likely contributed to the increasing prevalence of overweight in Indonesia. First, economic development, urbanisation and globalisation have altered the food environment in Indonesia,[@R3] as these factors result in easier access to and demand for processed foods.[@R4] In Indonesia, food availability per capita has increased by 40% over the last two decades, which is largely driven by an increase in the availability of fats (eg, palm oil).[@R6] Subsequently, poor dietary habits are now common among Indonesians. Data from the 2013 National Basic Health Research Survey indicated inadequate consumption of fruits and vegetables among the majority of Indonesians, and one prior study reported that higher consumption of meat and dairy was associated with a higher prevalence of obesity in Indonesia.[@R7] Prior evidence from Indonesia and the Southeast Asia region also suggests widespread consumption of ultra-processed foods and a positive association between consumption of processed foods, meat, dairy and 'Western foods' and overweight prevalence among children and adults.[@R6] Second, economic development and urbanisation also result in decreased physical activity and the Indonesian population is increasingly adopting a more sedentary lifestyle.[@R9] In part, this is due to changes in technology, which have led to more mechanised agricultural production and shifts away from agricultural-based employment. In addition, Indonesians perceive that increased motorised transport and rapid changes in the built environment have resulted in reduced physical activity.[@R13] Limited data also suggest that there are few bike lanes, sidewalks or parks in Indonesia.[@R6]
The objectives of this study were twofold. First, we document the changes in overweight prevalence that have occurred in Indonesia, between 1993 and 2014, among adults (aged ≥19 years), adolescents (aged 13--18 years), school-aged children (aged 6--12 years) and young children (aged 0--5 years). Second, this study provides a comprehensive examination of risk factors for overweight, by age, using regression-based methods. We believe the evidence will be useful for future advocacy efforts on obesity prevention and to inform relevant policy dialogues and programme design.
Methods {#s2}
=======
Study context {#s2-1}
-------------
Indonesia is the world's fourth most populous country, home to approximately 260 million individuals, and is the largest economy in Southeast Asia.[@R14] Indonesia's gross domestic product per capita has steadily risen, from \$807 in 2000 to \$3877 in 2018.[@R15] Over the last 40 years, Indonesia has experienced a process of rapid urbanisation and industrialisation, which have contributed to Indonesia's changing infrastructure. More than half of the population now live in an urban area and urbanisation is increasing at a rate of 2.3% annually. Economic development and urbanisation are related to changes in diet and physical activity. At the same time, the demographic transition (ie, the shift from a pattern of high fertility and high mortality to one of low fertility and low mortality) affects and is affected by nutritional change. Indonesia is in the middle phase of their demographic transition, expecting to harness the peak of demographic dividend between 2020 and 2030.[@R16]
Survey design and study population {#s2-2}
----------------------------------
These analyses used the publicly available Indonesian Family Life Survey (IFLS), which has previously been detailed elsewhere.[@R17] The IFLS is a longitudinal, home-based survey that collected socioeconomic and health data on individual respondents including children aged 0--18 and adults ≥19 years, and their households over time. The survey was first fielded in 1993 and there have been four subsequent rounds of data collection (1997, 2000, 2007, 2014). Follow-up surveys were fielded on the full sample and split-off households. Therefore, new members were added to the panel during each subsequent survey wave.
As described in detail elsewhere,[@R17] the original, multistage sampling frame was based on households from 13 out of 27 Indonesian provinces: North Sumatra, West Sumatra, South Sumatra, Lampung, DKI Jakarta, West Java, Central Java, Yogyakarta, East Java, Bali, West Nusa Tenggara, South Kalimantan and South Sulawesi. These provinces were selected to maximise representation of the population, capture the cultural and socioeconomic diversity of Indonesia and be cost-effective. The sample represented approximately 83% of the Indonesian population in 1993.[@R17] Within each of the 13 provinces, 321 enumeration areas were randomly chosen from the nationally representative sample frame used in the 1993 National Socioeconomic Survey. Urban areas were oversampled in the original survey; 20 households were randomly selected from each urban enumeration area and 30 households were randomly selected from each rural enumeration area. Original and split-off households were recontacted in subsequent survey waves. Indonesia now has 34 provinces, as eight have been added since 1999. As individuals moved between provinces within Indonesia over time, additional provinces were represented in the survey. In 2014, 24 provinces were represented.
The IFLS was approved by institutional review boards in the USA at RAND Corporation and in Indonesia at the University of Gadjah Mada. Survey participants provided written informed consent. A parent or guardian (typically the mother) provided informed consent for children younger than 11 years.[@R19] The data used for this study are retrospective and the authors did not have access to any identifying information. As such, ethical approval was not required.
Survey questions {#s2-3}
----------------
Many questions were consistently asked during each survey wave, but some were added or eliminated over the 20-year survey period to reflect the changing economic context. The household questionnaire collected information on household characteristics such as its size, physical infrastructure, access to sanitation and water sources, assets and food expenditures (see [online supplementary table 1](#SP1){ref-type="supplementary-material"}). Each adult respondent was asked about characteristics related to demographics (eg, marital status, age), socioeconomic status (eg, educational attainment, occupation) and health history (eg, smoking). Beginning in 2000, adults were also asked about their food intake in the past week; specifically, respondents indicated whether they ate each food type (eg, in the last week, did you eat any \[...\]) and the frequency of consumption (eg, how many days in a week did you eat \[...\] in the last week). In 2000 and 2007, self-reported diet questions focused on staple foods (eg, rice, eggs, dairy) and iron- and vitamin-A rich foods (eg, meat, green leafy vegetables, carrots, sweet potatoes). Ultra-processed foods (instant noodles, fast food, soft drinks, fried snacks) were added in 2014. In addition, married women were asked about their reproductive history and breast feeding and complementary feeding practices for children born within the 5 years prior to the survey.
10.1136/bmjopen-2019-031198.supp1
In 2007 and 2014, adults were asked about the type, duration (eg,\<2 hours or ≥2 hours) and number of days of physical activities they engaged during the last week. Activity type included vigorous activity, moderate activity and walking, using a modified version of the International Survey on Physical Activities. Vigorous physical activity was defined as any activities that make you breathe much harder than normal (eg, heavy lifting, digging, cycling with loads). Moderate physical activity was defined as any activities that make you breathe somewhat harder than normal (eg, carrying light loads, bicycling or mopping the floor). Walking included at work and at home, walking from place to place and walking for recreation.
The child questionnaire paralleled the adult questionnaire, with age-appropriate modifications. Specifically, information was collected on age, gender, educational attainment, morbidities and food frequency. Physical activity was not collected among children. For children younger than 11 years, the mother, female guardian or household caretaker answered the questions. Children aged ≥11 years were allowed to respond for themselves.
Height and weight were measured for adults and children by trained interviewers, using a rigorous research protocol.[@R17] Interviewers completed both didactic and hands-on anthropometric training and after completing each questionnaire and measurements in the field, interviewers checked for completeness and errors, using a computer-assisted programme. Supervisors randomly observed interviewers to ensure that measurements were being performed according to the training instructions. Height was measured to the nearest millimetre using a Seca plastic height board. Recumbent height was measured among children aged \<24 months. Weight was measured to the nearest one-tenth of a kilogram using a Camry model EB1003 scale.
Statistical analysis {#s2-4}
--------------------
Overweight served as our a priori primary endpoint. Among adults (aged ≥19 years), overweight was defined as body mass index (BMI) ≥25 kg/m^2^. Among children and adolescents aged 6--18 years, overweight was defined as BMI-for-age z-score \>1 using the WHO reference.[@R20] Among children aged 0--5 years, overweight was defined as weight-for-height z-score \>2 using the Multicentre Growth Reference Study Standard.[@R21]
Appropriate cut-offs were applied to create dichotomous or categorical variables for age, educational attainment and occupation. Urban/rural residence, diet and physical activity data were modelled as binary variables. Household wealth quintiles were created using principal component analysis, using following variables: type of floor material, type of toilet, type of cooking fuel and ownership of assets including: land, livestock, vehicle(s), household appliances, furniture and utensils, jewellery and monetary savings. Dichotomous variables were created from these categorical variables and a frequency test was run for each of the new variables to ensure that those variables with a variance of zero were removed. The wealth score was then split into quintiles based on the distribution of the data.
Household food expenditures on rice and cooking oil represented the household-level expenditure on each item as a percentage of the households' total expenditures on food. Food expenditures were then dichotomised as lower and higher based on the mean value in each survey year. Rice was selected as it is a key staple food in Indonesia. Cooking oil was also selected because it has been shown to perpetuate energy imbalance.
Descriptive statistics are presented as percentages for categorical variables or means for continuous variables. The distribution of overweight by gender (adult women and adult men aged ≥19 years) and age (0--5 years, 6--12 years, 13--18 years) was detailed for each survey year. Then, we described individuals who were overweight, by the aforementioned characteristics (eg, education, wealth), from 1993 to 2014. Finally, we described the distribution of the consumption of ultra-processed foods and of physical activity among adults in 2014.
Prior conceptual models,[@R2] along with the availability of data in the IFLS were used to identify plausible determinants of the obesity endemic in Indonesia. To create a parsimonious model, univariate analyses first explored risk factors and their association with overweight, using cross-sectional logistic regression models, for five groups in 2014: adult women, adult men, children aged 0--5 years, children aged 6--12 years and children aged 13--18 years. Largely, variables were included in multivariable models if they were statistically significantly associated with the dependent variable in univariate models (defined as p\<0.05). Age, gender, urban/rural residence and wealth were included in the multivariable logistic regression models irrespective of statistical significance. Pregnant women were excluded from analyses. We assessed multicollinearity in all models.
In sensitivity analyses, we explored the association between selected risk factors and overweight among adults, over time (1993--2014), using modified Poisson models.[@R24] These longitudinal models were considered sensitivity analyses because individual-level consumption of ultra-processed foods and physical activity data were not available in prior years. Alpha was set to 0.05 to determine statistical significance. All analyses employed sampling weights, which account for the survey design and were performed using Stata 15.1 (StataCorp, College Station, Texas, USA).
Patient and public involvement {#s2-5}
------------------------------
Neither patients nor the public were involved in this secondary data analysis.
Results {#s3}
=======
Between 1993 and 2014, the prevalence of overweight among adults doubled from 17.1% to 33.0% ([figure 1](#F1){ref-type="fig"}). The prevalence of overweight among children under 5 years increased from 4.2% to 9.4% between 1993 and 2007, but since then, remained relatively stagnant. Among children aged 6--12 years and 13--18 years, the prevalence of overweight increased from 5.1% to 15.6% and from 7.1% to 14.1% between 1993 and 2014, respectively.
{#F1}
The characteristics of overweight individuals are presented in [tables 1--4](#T1 T2 T3 T4){ref-type="table"} and in [online supplementary tables 2--4](#SP1){ref-type="supplementary-material"}. More women and men over 40 years old (vs younger ages) were overweight over the study period ([tables 1 and 2](#T1 T2){ref-type="table"}). Most overweight women and men had at least a primary school education, and more than half lived in urban areas. The distribution of wealth among overweight adults varied over time. In 1993, the largest percentage of overweight women (29.4%) was in the wealthiest households, whereas in 2014, wealth was evenly distributed (\~20%). Trends were similar for men.
######
Per cent distribution of selected characteristics among overweight adult women aged 19+ years in Indonesia over time (1993-2014)
1993\*† 1997\*† 2000\*† 2007\*† 2014\*†
----------------------- --------- --------- --------- --------- --------- ------ ------ ------ ------ ------
Age (years)
19--29 97 18.9 248 18.7 423 14.7 848 15.6 1197 10.9
30--39 253 38.1 531 32.6 728 30.5 1167 24.8 1937 19.6
40--49 178 25.4 473 27.0 699 29.9 1044 31.1 1511 31.9
50--59 97 14.3 255 13.2 342 15.4 618 17.6 996 24.4
≥60 24 3.3 169 8.6 254 9.6 389 10.9 551 13.2
Education
No education 39 7.3 225 14.5 276 11.2 279 9.1 263 6.9
Primary 415 66.7 832 61.7 1232 54.6 1702 53.7 2198 50.4
Junior or senior 152 20.3 308 19.4 711 27.0 1109 28.6 2077 32.8
University 43 5.7 69 4.4 172 7.2 335 8.6 623 9.9
Parity‡
≤2 16 15.2 20 43.1 71 20.1 56 20.1 428 35.8
3--4 19 17.7 8 14.2 72 18.4 76 24.9 230 29.9
≥5 84 67.1 20 42.7 254 61.5 191 55.0 306 34.4
Smoking status
Non-smoker 614 96.7 1553 95.6 2311 96.1 3889 97.5 6027 97.4
Smoker 25 3.3 73 4.4 99 3.9 101 2.5 134 2.6
Marital status
Never married 10 2.8 85 5.0 146 4.8 237 4.9 291 3.1
Married 582 89.1 1336 82.7 1926 80.4 3298 81.9 5185 81.7
Other 49 8.1 220 12.3 351 14.8 475 13.2 714 15.2
Employment
Not working 362 57.0 803 50.8 969 39.8 1527 37.6 2178 35.2
Agriculture 27 5.5 135 8.8 260 11.6 517 14.5 816 15.6
Skilled manual§ 44 6.2 141 9.3 201 9.3 315 8.2 479 8.4
Skilled¶ 202 31.3 545 31.1 956 39.2 1616 39.7 2618 40.8
Residence
Rural 150 33.4 583 43.0 893 41.5 1604 47.3 2286 46.0
Urban 499 66.6 1093 57.0 1553 58.5 2462 52.7 3906 54.0
Wealth
Lowest 81 14.4 238 17.3 323 15.2 603 16.3 1095 23.0
Second 64 10.3 233 16.9 388 19.5 875 27.6 1087 22.7
Middle 163 21.0 312 23.4 430 20.9 521 15.8 781 17.3
Fourth 203 24.9 270 19.9 426 20.8 679 20.4 894 18.7
Highest 645 29.4 330 22.5 475 23.5 645 19.9 848 18.3
Family size
≤4 249 40.2 533 33.3 1161 48.8 1274 30.2 3898 65.3
\>4 400 59.8 1143 66.7 1285 51.2 2792 69.8 2294 34.7
Province
Bali 78 18.6 213 13.0 305 14.0 522 16.8 774 17.0
Central Java 107 26.0 243 22.7 369 24.2 576 24.3 885 25.1
West Java 92 20.4 272 23.5 416 25.6 606 19.4 955 19.8
Other 372 36.7 948 46.8 1353 40.2 2171 42.1 3232 40.8
Food expenditures\*\*
*Rice*
Lowest 186 49.5 880 53.1 1242 51.6 2114 53.2 3137 53.0
Highest 205 50.5 790 46.9 1198 48.4 1951 46.8 3051 47.0
*Cooking oil*
Lowest 198 42.5 786 49.3 1174 48.6 1827 45.4 3562 58.6
Highest 340 57.5 882 50.7 1266 51.4 2238 54.6 2625 41.4
\*Overweight is defined as body mass index ≥25 kg/m^2^. Percentages are weighted to be representative of the population of Indonesia in the given survey year.
†Excludes women who are currently pregnant.
‡Parity is only queried among ever-married women.
§Skilled manual labour combines the following employment sectors: mining, manufacturing, electric, gas, water maintenance and construction.
¶Skilled labour combines the following employment sectors: retail and service, transportation.
\*\*Represents the household level expenditure on each item as a percentage of the households' total expenditures on food.
######
Per cent distribution of selected characteristics among overweight men aged 19+ years in Indonesia over time (1993-2014)
1993\* 1997\* 2000\* 2007\* 2014\*
-------------------- -------- -------- -------- -------- -------- ------ ------ ------ ------ ------
Age (years)
19--29 27 7.3 88 13.4 224 14.1 419 15.6 573 16.5
30--39 117 34.3 196 30.7 356 28.7 698 24.6 1048 21.7
40--49 121 33.4 215 32.0 336 31.8 549 30.6 885 25.4
50--59 52 14.4 123 16.1 184 17.2 326 20.7 514 25.0
≥60 35 10.6 62 7.8 99 8.2 164 8.5 262 11.4
Education
No education 6 2.0 24 3.8 32 2.5 32 1.8 39 1.8
Primary 153 44.6 227 40.8 385 34.2 553 34.5 725 32.4
Junior or senior 124 37.9 210 35.6 519 43.0 805 42.6 1288 45.0
University 68 15.5 112 19.8 247 20.3 386 21.1 612 20.8
Smoking status
Non-smoker 163 44.9 289 41.3 527 44.8 897 43.6 1481 46.5
Smoker 181 55.1 371 58.7 651 55.2 1212 56.4 1779 53.5
Marital status
Never married 4 3.0 54 8.2 131 8.6 218 8.9 353 9.6
Married 342 96.8 605 90.4 1031 89.7 1850 88.2 2833 87.1
Other 1 0.2 11 1.4 22 1.7 56 2.9 95 3.3
Employment
Not working 34 12.4 70 9.8 84 7.4 165 8.7 248 8.7
Agriculture 27 9.9 100 15.2 179 15.8 328 16.1 500 16.7
Skilled manual† 87 25.8 123 19.7 199 17.5 377 17.7 598 18.8
Skilled labour‡ 185 51.9 366 55.4 687 59.4 1231 57.6 1874 55.9
Residence
Rural 73 33.5 211 35.7 381 36.0 717 39.0 987 37.4
Urban 279 66.5 473 64.3 818 64.0 1439 61.0 2294 62.6
Wealth
Lowest 22 7.3 82 11.8 130 10.9 324 15.6 547 19.9
Second 26 6.7 88 14.1 166 16.1 460 23.9 562 20.1
Middle 61 18.7 135 22.5 214 19.4 304 15.5 445 17.0
Fourth 72 20.6 118 19.7 267 23.8 418 22.1 554 19.9
Highest 169 46.6 205 31.9 315 29.8 418 22.9 661 23.1
Family size
≤4 128 36.9 218 32.7 579 50.0 712 30.9 2061 63.9
\>4 224 63.1 466 67.3 620 50.0 1444 69.1 1220 36.1
Province
Bali 38 16.5 70 9.7 116 10.8 227 14.4 369 14.0
Central Java 58 29.6 88 20.3 162 22.6 289 23.2 457 24.2
West Java 38 17.2 100 23.2 195 26.4 308 20.3 433 21.0
Other 141 36.7 313 46.8 540 40.2 980 42.1 1531 40.8
Food expenditures§
*Rice*
Lowest 83 43.0 365 53.5 627 52.0 1135 53.5 1709 54.5
Highest 103 57.0 314 46.5 564 48.0 1016 46.5 1561 45.5
*Cooking oil*
Lowest 89 34.5 290 43.6 525 45.2 936 42.8 1811 56.6
Highest 203 65.5 388 56.4 665 54.8 1215 57.2 1459 43.4
\*Overweight is defined as body mass index ≥25 kg/m^2^. Percentages are weighted to be representative of the population of Indonesia in the given survey year.
†Skilled manual labour combines the following employment sectors: mining, manufacturing, electric, gas, water maintenance and construction.
‡Skilled labour combines the following employment sectors: retail and service, transportation.
§Represents the household level expenditure on each item as a percentage of the households' total expenditures on food.
######
Distribution and frequency of consumption of ultra-processed foods and physical activity in Indonesia, 2014
Women Men Children aged 6 months--5 years Children aged 6--12 years Children aged 13--18 years
--------------------- -------- ------ --------------------------------- --------------------------- ---------------------------- ------ ------ ------ ------ ------
Instant noodles
Last week\* 9818 60.0 8703 63.7 2674 56.8 5467 77.5 2084 78.3
Mean days† 2.5 2.6 2.7 3.0 2.9
Fast food
Last week\* 1718 9.0 1291 8.8 685 14.5 1357 18.8 450 16.3
Mean days† 1.8 1.7 2.4 2.5 2.1
Soda
Last week\* 1956 10.7 3281 22.6 278 5.0 1075 14.5 647 23.0
Mean days† 1.8 2.0 2.0 2.1 2.0
Fried snacks
Last week\* 9706 65.4 8668 68.8 2282 50.4 4530 66.4 1841 72.0
Mean days† 4.0 3.9 3.3 3.9 4.1
Vigorous activity‡
Last week 15 191 10.7 13 148 38.0 na na na
Mean days 3.9 4.2 na na na
Moderate activity‡
Last week 15 191 58.1 13 148 54.2 na na na
Mean days 4.8 4.4 na na na
Walking‡
Last week 15 191 68.8 13 148 72.4 na na na
Mean days 5.1 5.1 5.1 na na na
\*Per cent distribution, weighted to be representative of the population of Indonesia in the given survey year. n represents the number of respondents that consumed a particular item during the last week.
†The average number of days an individual consumed each food is queried only if the respondent reported that they consumed the item in the last week.
‡Defined using the International Physical Activity Questionnaire. Respondents are asked about each type of physical activity (ie, vigorous, moderate, walking) separately and therefore, these should be interpreted as mutually exclusive categories. The per cent distribution is weighted to be representative of the population of Indonesia in the given survey year.
na, not applicable.
######
Multivariable logistic regression investigating the association between selected characteristics and overweight among adults aged 19+ years in Indonesia, 2014
Women†‡ Men†‡
-------------------------------------------------- --------- ----------------------- ------ -----------------------
Age (years)
19--29 2403 Reference 1441 Reference
30--39 2630 2.03 (1.75 to 2.34)\* 1807 1.79 (1.40 to 2.30)\*
40--49 1943 2.83 (2.41 to 3.31)\* 1320 2.37 (1.83 to 3.08)\*
50--59 1357 2.36 (1.97 to 2.83)\* 719 2.37 (1.76 to 3.20)\*
60+ 740 1.68 (1.34 to 2.12)\* 488 0.92 (0.62 to 1.36)
Education
No education 511 Reference 106 Reference
Primary 3562 2.04 (1.59 to 2.60)\* 2029 1.76 (0.87 to 3.57)
Junior or senior 3809 1.89 (1.46 to 2.46)\* 2854 2.47 (1.21 to 5.04)\*
University 1191 1.67 (1.24 to 2.26)\* 786 3.66 (1.75 to 7.65)\*
Smoking status
Non-smoker 8881 Reference 1849 Reference
Smoker 192 0.78 (0.55 to 1.10) 3926 0.51 (0.43 to 0.60)\*
Marital status
Never married 683 Reference 774 Reference
Married 7909 2.39 (1.88 to 2.98)\* 4931 2.04 (1.48 to 2.82)\*
Other 481 1.54 (1.10 to 2.15)\* 70 1.59 (0.77 to 3.27)
Employment
Not working 3227 Reference 371 Reference
Agriculture 1585 0.66 (0.56 to 0.78)\* 1483 0.50 (0.35 to 0.73)\*
Skilled manual§ 794 0.92 (0.76 to 1.12) 1290 0.73 (0.51 to 1.06)
Skilled¶ 3467 1.17 (1.04 to 1.33)\* 2631 1.15 (0.82 to 1.61)
Physical activity\*\*:
No vigorous activity na 3624 Reference
Vigorous activity na 2151 0.93 (0.79 to 1.10)
No moderate activity 3663 Reference na
Moderate activity 5410 1.10 (0.99 to 1.22) na
Consumed last week:
*Instant noodles*
No 3218 Reference na
Yes 5855 1.10 (0.98 to 1.25) na
*Fast food*
No 8116 Reference 5172 Reference
Yes 957 1.02 (0.85 to 1.27) 603 1.34 (1.06 to 1.69)\*
*Soda*
No na 4196 Reference
Yes na 1579 1.10 (0.92 to 1.30)
*Fried snacks*
No 3236 Reference 1782 Reference
Yes 5837 1.12 (1.00 to 1.25)\* 3993 1.06 (0.89 to 1.25)
Mean number of days consumed in the last week††:
Instant noodles na 5775 0.99 (0.95 to 1.04)
Household level
Food expenditures‡‡
*Rice*
Lowest 4408 Reference na
Highest 4665 1.02 (0.91 to 1.13) na
*Cooking oil*
Lowest 5256 Reference 3378 Reference
Highest 3817 1.24 (1.12 to 1.38)\* 2397 1.19 (1.02 to 1.39)\*
Residence
Rural 3803 Reference 2401 Reference
Urban 5,270 1.21 (1.08 to 1.35)\* 3374 1.26 (1.06 to 1.50)\*
Wealth
Lowest 2289 Reference 1354 Reference
Second 2017 1.23 (1.06 to 1.44)\* 1271 0.95 (0.76 to 1.20)
Middle 1455 1.23 (1.03 to 1.45)\* 964 0.99 (0.77 to 1.26)
Fourth 1696 1.06 (0.90 to 1.25) 1127 1.06 (0.84 to 1.34)
Highest 1616 1.12 (0.95 to 1.32) 1059 1.17 (0.92 to 1.48)
\*p\<0.05.
†Overweight is defined as body mass index ≥25 kg/m^2^.
‡ORs and CIs are estimated using logistic regression and are weighted to account for the survey design. Pregnant women are excluded.
§Skilled manual labour combines the following employment sectors: mining, manufacturing, electric, gas, water maintenance and construction.
¶Skilled labour combines the following employment sectors: retail and service, transportation.
\*\*Defined using the International Physical Activity Questionnaire.
††Modelled as a continuous variable, the average number of days consumed is queried if the respondent reported that they consumed item during the last week.
‡‡Indicates the household level expenditure on each item as a percentage of the households' total expenditures on food.
na, not applicable, not included in multivariable model.
Nearly equal proportions of boys and girls aged 0--19 years were overweight and a majority had a parent with at least primary school education ([online supplementary tables 2--4](#SP1){ref-type="supplementary-material"}). Across all groups, the proportion of overweight children who had an overweight mother increased between 1993 and 2014. The distribution of urban/rural residence among overweight children also changed over time, with more overweight young and school-aged children residing in urban areas by 2014 (\~60%) compared with 1993. The distribution of wealth among overweight children and adolescents also varied over time, but like adults, by 2014, there was an even distribution of overweight children in each wealth quintile (\~20%).
Distribution of consumption of ultra-processed food and physical activity {#s3-1}
-------------------------------------------------------------------------
In 2014, approximately 60% of adults reported consuming instant noodles during the last week ([table 3](#T3){ref-type="table"}). Likewise, about 65% of adults reported consuming fried snacks, for 4.0 days on average. Smaller proportions of adults regularly consumed fast food or soda.
Trends were similar among children. Nearly 60% of children aged 6--59 months consumed instant noodles during the last week, as did three-quarters of children aged 6--12 years (77.5%) and 13--18 years (78.3%). Half (50.4%) of children aged 6--59 months and about two-thirds of children aged 6--18 years regularly consumed fried snacks. Similar proportions of children regularly consumed fast food (\~15%), irrespective of age. All age groups of children consumed soda an average of 2 days per week.
Higher proportions of men were engaged in vigorous physical activity (38.0%) compared with women (10.7%). More than half adults reported being engaged in moderate physical activity and approximately 70% reported walking during the last week. Physical activity levels decreased, among both women and men, between 2007 and 2014 (2007 data not shown).
Regression-based analyses among adults {#s3-2}
--------------------------------------
[Figure 2](#F2){ref-type="fig"} summarises all associations between hypothesised risk factors and overweight in multivariable models (see [online supplementary tables 5 and 6](#SP1){ref-type="supplementary-material"} for univariable models). Among women, older age, higher education, living in an urban area and higher wealth were associated with higher odds of overweight in 2014 ([table 4](#T4){ref-type="table"}). Married women versus never married also had higher odds of overweight (OR=2.39; 95% CI: 1.88 to 2.98). Working in a skilled labour position, versus not working, was associated with 17% higher odds of overweight among women (OR=1.17; 95% CI: 1.04 to 1.33), whereas working in agriculture-based labour, a physically demanding occupation, was associated with an approximately 30% lower odds of overweight (OR=0.66; 95% CI: 0.56 to 0.78). Household expenditures on cooking oil (OR=1.24; 95% CI: 1.12 to 1.38) and individual's consumption of fried snacks (vs no consumption) (OR=1.12; 95% CI: 1.00 to 1.25) were associated with higher odds of overweight.
{#F2}
Similarly, among men, older age, higher educational attainment and living in an urban area were associated with higher overweight ([table 4](#T4){ref-type="table"}). Wealth was not associated with overweight among men. Married men versus never married had twofold higher odds of overweight (OR=2.04; 95% CI: 1.48 to 2.82). Consuming fast food, compared with not consuming fast food, was associated with 34% higher odds of overweight among men (OR=1.34; 95% CI: 1.06 to 1.69). Current smokers (vs non-smokers) (OR=0.51; 95% CI: 0.43 to 0.60) and men who had an agriculture-based occupation (vs not working) (OR=0.50; 95% CI: 0.35 to 0.73) had 50% lower odds of overweight. Results were similar when investigating associations among adults in longitudinal models ([online supplementary tables 7 and 8](#SP1){ref-type="supplementary-material"}).
Regression-based analyses among children {#s3-3}
----------------------------------------
In multivariable models for children aged 6 months to 5 years, consuming fast food (OR=1.48; 95% CI: 1.00 to 2.19), urban residence (OR=1.64; 95% CI: 1.14 to 2.35) and higher household wealth (OR=1.70; 95% CI: 1.03 to 2.81) were associated with higher odds of overweight ([table 5](#T5){ref-type="table"}). Among children aged 6--12 years, every 1 year increase in age was associated with 6% higher odds of overweight (OR=1.06; 95% CI: 1.01 to 1.11). Having an overweight mother (OR=1.92; 95% CI: 1.59 to 2.32), living in an urban area (OR=1.55; 95% CI: 1.26 to 1.91) and higher household wealth were associated with higher odds of overweight. On the contrary, parental employment in agriculture-based (OR=0.61; 95% CI: 0.44 to 0.85) or skilled manual labour (OR=0.50; 95% CI: 0.33 to 0.74) and living in households with more than four people (OR=0.82; 95% CI: 0.68 to 0.99) were associated with lower odds of overweight.
######
Multivariable logistic regression models investigating the association between selected characteristics and overweight among children in Indonesia, 2014
6 months--5 years† 6--12 years‡ 13--18 years‡
---------------------- -------------------- ----------------------- --------------- ----------------------- ------ -----------------------
Mean age§ 3424 1.01 (0.88 to 1.16) 4791 1.06 (1.01 to 1.11)\* 3186 0.84 (0.79 to 0.91)\*
Gender
Boys 1767 Reference 2455 Reference 1648 Reference
Girls 1657 0.78 (0.58 to 1.08) 2336 0.96 (0.79 to 1.15) 1538 1.06 (0.85 to 1.33)
Parental education¶
No education 188 Reference 313 Reference 264 Reference
Primary 1234 0.97 (0.40 to 2.36) 1980 0.73 (0.48 to 1.11) 1505 1.01 (0.62 to 1.66)
Junior or senior 1568 1.12 (0.47 to 2.65) 1997 0.84 (0.55 to 1.27) 1173 1.18 (0.71 to 1.94)
University 434 1.12 (0.44 to 2.87) 501 1.34 (0.84 to 2.14) 244 1.38 (0.76 to 2.49)
Parental employment¶
Not working 1780 Reference 1989 Reference 1239 Reference
Agriculture 433 0.61 (0.31 to 1.21) 821 0.61 (0.44 to 0.85)\* 581 0.64 (0.43 to 0.96)\*
Skilled manual\*\* 208 0.98 (0.52 to 1.84) 371 0.50 (0.33 to 0.74)\* 264 0.52 (0.33 to 0.83)\*
Skilled†† 1003 1.15 (0.82 to 1.63) 1610 0.91 (0.73 to 1.13) 1102 0.70 (0.53 to 0.91)\*
Maternal overweight
Not overweight 2238 Reference 2894 Reference 1834 Reference
Overweight 1186 1.20 (0.89 to 1.62) 1897 1.92 (1.59 to 2.32)\* 1352 1.92 (1.53 to 2.43)\*
Consumed last week:
*Instant noodles*
No 1525 Reference na na
Yes 1899 0.79 (0.57 to 1.11) na na
*Fast food*
No 2952 Reference 3909 Reference na
Yes 472 1.48 (1.00 to 2.19)\* 882 1.16 (0.93 to 1.45) na
Residence
Rural 1478 Reference 2053 Reference 1323 Reference
Urban 1946 1.64 (1.14 to 2.35)\* 2738 1.55 (1.26 to 1.91)\* 1863 1.74 (1.33 to 2.27)\*
Wealth
Lowest 830 Reference‡‡ 1109 Reference 799 Reference
Second 807 1.54 (0.94 to 2.53) 1056 1.43 (1.06 to 1.92)\* 735 1.11 (0.80 to 1.54)
Middle 565 1.43 (0.83 to 2.48) 812 1.73 (1.26 to 2.36)\* 513 0.95 (0.66 to 1.37)
Fourth 671 1.70 (1.03 to 2.81)\* 931 1.45 (1.07 to 1.95)\* 574 0.95 (0.65 to 1.37)
Highest 551 1.61 (0.96 to 2.70) 883 1.55 (1.14 to 2.11)\* 565 0.94 (0.66 to 1.33)
Family size
≤4 na 2105 Reference 1225 Reference
\>4 na 2686 0.82 (0.68 to 0.99)\* 1961 0.87 (0.69 to 1.10)
\*p\<0.05.
†Overweight is defined as weight-for-height z-score \>2 using the MGRS/WHO Standard. Coefficients are estimated using logistic regression models and are weighted to account for the survey design.
‡Overweight is defined body mass index z-score \>1 using the WHO reference. Coefficients are estimated using logistic regression models and weighted to account for the survey design.
§Modelled as a continuous variable.
¶Parental characteristics are of the mother, when available. When not available, the characteristic is based on the father's data.
\*\*Skilled manual labour combines the following employment sectors: mining, manufacturing, electric, gas, water maintenance and construction.
††Skilled labor combines the following employment sectors: retail and service, transportation.
‡‡Wealth was not statistically significant in the univariable model but is included in the multivariable model due to previously established associations.
MGRS, Multicentre Growth Reference Study; na, not applicable, not included in multivariable model.
Having an overweight mother was associated with 92% higher odds of being overweight among adolescents (OR=1.92; 95% CI: 1.53 to 2.43). Living in an urban area also predicted higher odds of overweight among this age group (OR=1.74; 95% CI: 1.33 to 2.27). On the contrary, having a parent employed in agriculture-based labour (OR=0.64; 95% CI: 0.43 to 0.96), skilled manual labour (OR=0.52; 95% CI: 0.33 to 0.83) or skilled labour (OR=0.70; 95% CI: 0.53 to 0.91), compared with not working was associated with lower odds of overweight among adolescents.
Discussion {#s4}
==========
Several aspects of our results paint an alarming picture of obesity-related health in Indonesia. Increases in overweight prevalence were observed among all groups. Between 1993 and 2014, the prevalence of overweight increased by 100% and 88%, among women and men, respectively. Similarly, among school-aged children and adolescents, the prevalence of overweight increased by more than 100% between 1993 and 2014. Over the past two decades, more rural than urban residents became overweight, and in 2014, the proportions of overweight adults were evenly distributed among across wealth quintiles, suggesting that overweight is now prevalent among the poor. Moreover, consuming instant noodles (high in sodium) and fried snacks (high in fat) are now regularly consumed among adults and children in Indonesia and consuming fast food and fried snacks were associated with higher odds of overweight. At the same time, the rate of overweight has remained stagnant among young children, between 2007 and 2014, and being employed in a physically demanding job (ie, agriculture-based labour) remains protective against overweight.
More than half of Indonesians now live in an urban area and living in an urban area was associated with higher odds of overweight. Several other analyses have also reported higher rates of overweight in urban areas in lower-income countries[@R25] and in Indonesia specifically.[@R27] Urbanisation alters supply-side factors, such as the food system and the rapid spread of supermarkets in low/middle-income countries (LMICs), which makes low-cost, convenient and highly processed foods more available and accessible.[@R31] Simulations suggest that a shift in the urban population from 25% to 75% is associated with a 4-percentage point increase of total energy from fat and an additional 12-percentage points of energy from sugar.[@R34] In addition, urbanisation leads to lower physical activity, in part because the jobs that become available tend to be more sedentary.
Paralleling increased country-level economic development, the within-country socioeconomic context has changed in Indonesia. Poverty has dramatically decreased over the last 20 years, while educational attainment has increased; in 2016, approximately 34% of Indonesians had at least completed upper secondary school, compared with only 26% in 2006.[@R14] Relatedly, we found that higher education was associated with higher odds of overweight among both men and women. Higher wealth was also associated with 25% higher odds of being overweight among women and a 70%, 55% and 74% higher odds of overweight among young children, school-age children and adolescents, respectively. Higher education[@R7] and higher wealth or income[@R7] have been shown to be associated with higher overweight among adults and children in prior studies conducted in Indonesia. But importantly, descriptive statistics in our study also show a growing prevalence of overweight among lower-wealth groups. For example, the per cent of overweight women in the poorest and second wealth quintiles grew by 60% and 120%, respectively. This is consistent with prior studies, which have documented an increasing burden of overweight among women with lower education and in poorer households in Indonesia and other LMICs.[@R36]
For both adults and children, consuming ultra-processed foods (instant noodles, fried snacks, fast food) was associated with higher odds of overweight. Prior literature on the consumption of ultra-processed foods in Indonesia is limited. But these results are generally consistent with prior findings that reported that the per cent of total energy coming from fat has increased in this context[@R27] and one recent survey conducted in Asia, which suggested that the consumption of palm oil, 'Western' foods and processed foods are widespread.[@R6]
As Indonesia has continued to economically develop, more women are entering the labour force and less of the population is employed in agriculture-based labour. Currently, about 50% of Indonesia's labour force is employed in the service sector, compared with 30% in agriculture and 20% in industry.[@R14] Similar to prior literature, employment was associated with overweight in this sample.[@R41] We found that skilled occupations were associated with higher odds of overweight, whereas agriculture-based employment was associated with lower odds of overweight. This is likely related to occupation-related physical activity; Monda and colleagues provide empirical evidence of a reduction in the intensity of occupational activity with economic development in China.[@R43] In fact, reduced physical activity likely represents the most important cause of rapid overweight increases between 1990 and 2010 in most LMICs, with dietary factors playing the major role from 2010 to present.[@R44]
Similar to our results, Rachmi and colleagues report that overweight prevalence is higher among younger-aged boys versus girls in Indonesia.[@R7] Finally, having an overweight mother was strongly associated with higher odds of overweight among school-age children and adolescents. Although the importance of intergenerational nutrition is often a concern for undernutrition, it warrants consideration for overnutrition as well. Prepregnancy maternal overweight puts infants at increased risk of being born large-for-gestational age and subsequently overweight during childhood.[@R45] A child born to a mother who is overweight is twice as likely to be overweight as a child, compared with a child whose mother was normal weight.[@R48]
A recent report by the Economist Intelligence Unit reviewed obesity policies and interventions that are being/have been tested in Southeast Asian countries.[@R6] Interventions that targeted food intake showed the most promise to reduce obesity. Traditionally, nutrition programmes have focused on women and young children. But our findings call for urgent action to mitigate overweight among all age groups in Indonesia and highlight the need for prevention strategies that also target men and boys, as well as individuals in rural areas. A multipronged strategy would include approaches that aim to enhance the food environment, improve the food system and provide effective social behavioural change communication.[@R51] Fiscal measures, such as food-related and beverage-related taxes, to reduce the consumption of unhealthy items and subsidies for fruits and vegetables production and consumption, may be effective in improving the overall food environment and are a feasible approach for targeting most segments of the population in Indonesia.[@R52] Regulatory measures to control warning labelling systems and marketing of unhealthy foods and beverages may also improve the obesogenic environment.[@R52] It is also important to address overweight early, as child and adolescent obesity tracks into adulthood[@R58]; interventions aimed at improving nutrition literary and eating and physical activity behaviours through school-based nutrition education and effective social and behavioural change communication to mobilise the support of the entire community also warrant consideration.[@R52] Using technology-based platforms, such as social media, to deliver obesity-prevention messages may be a promising strategy in Indonesia given that population is quite young. A multipronged strategy that uses a number of delivery platforms (eg, policy, healthcare system, schools) is needed to mitigate the obesity epidemic among all age groups in Indonesia.
These findings and recommendations should be interpreted while keeping in mind the limitations of our study. First, we largely employ cross-sectional data, therefore, we cannot infer any causal associations and there may be unmeasured confounding. Data on the consumption of a select number of ultra-processed foods were only collected in 2014, which limits our ability to make inferences about the association between ultra-processed foods and overweight over time. Physical activity levels were based on respondent recall. These limitations in the IFLS and the potential for unmeasured confounding with a cross-sectional study design may have attenuated some associations related to food consumption and physical activity. Finally, these findings may not be generalisable beyond Indonesia. Despite some limitations, we are able to compare overweight trends over time among all age groups, to explore the consumption of ultra-processed foods and physical activity and investigate hypothesised risk factors for overweight using regression-based methods in Indonesia.
Indonesia is undergoing nutrition transition, as evidenced by the increasing prevalence of overweight, the ubiquitous consumption of ultra-processed foods and decreasing levels of physical activity. These data suggest that urgent programme and policy action is needed to reduce and prevent overweight among all age groups. In addition, these data highlight the need for prevention strategies that also target males, as well as the increasing prevalence of overweight among rural and poorer populations. Multisectoral, multistakeholder actions and solutions are needed to improve the food environment in Indonesia.
Supplementary Material
======================
###### Reviewer comments
###### Author\'s manuscript
The authors wish to thank Paul Pronyk for his insightful comments on the manuscript.
**Contributors:** VMO, MM and JHR conceptualised the research question and jointly developed the analytic plan. MM acquired the publicly available data. VMO took the primary role in data analysis, interpretation of findings, and drafting of the manuscript. MM and JHR also provided critical input on the interpretation of the findings and provided input on all manuscript drafts. VMO, MM and JHR approved the final version of the manuscript.
**Funding:** The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
**Competing interests:** None declared.
**Patient consent for publication:** Not required.
**Provenance and peer review:** Not commissioned; externally peer reviewed.
**Data availability statement:** Data are publicly available and can be downloaded at <https://www.rand.org/well-being/social-and-behavioral-policy/data/FLS/IFLS.html>
|
#if ARDUINO < 100
#include <WProgram.h>
#else
#include <Arduino.h>
#endif
#include <stddef.h>
#include "RF24Audio.h"
#include "RF24.h"
#include <userConfig.h>
//******* General Variables ************************
volatile boolean buffEmpty[2] = {true,true}, whichBuff = false, a, lCntr=0, streaming = 0, transmitting = 0;
volatile byte buffCount = 0;
volatile byte pauseCntr = 0;
unsigned int intCount = 0;
byte txCmd[2] = {'r','R'};
byte buffer[2][buffSize+1];
char volMod = -1;
byte bitPos = 0, bytePos = 25;
byte bytH;
byte radioIdentifier;
#if defined (tenBit)
unsigned int sampl;
byte bytL;
#endif
unsigned long volTime = 0;
RF24 radi(0,0);
#if defined(__AVR_ATmega1280__) || defined(__AVR_ATmega2560__) || (__AVR_ATmega32U4__) || (__AVR_AT90USB646__) || defined(__AVR_AT90USB1286__) || (__AVR_ATmega128__) ||defined(__AVR_ATmega1281__)||defined(__AVR_ATmega2561__)
#define rampMega
#endif
const byte broadcastVal = 255; // The value for broadcasting to all nodes using the broadcast() command.
/*****************************************************************************************************************************/
/************************************************* General Section ***********************************************************/
RF24Audio::RF24Audio(RF24& _radio, byte radioNum): radio(_radio){
radi = radio;
radioIdentifier = radioNum;
}
void RF24Audio::begin(){
radio.begin();
//delay(500);
// Set the defined input pins as inputs with pullups high. See http://arduino.cc/en/Tutorial/InputPullupSerial
#if defined (ENABLE_LED)
pinMode(ledPin,OUTPUT);
#endif
pinMode(speakerPin,OUTPUT); pinMode(speakerPin2,OUTPUT);
pinMode(TX_PIN,INPUT_PULLUP); pinMode(VOL_UP_PIN,INPUT_PULLUP);
pinMode(VOL_DN_PIN,INPUT_PULLUP); pinMode(REMOTE_TX_PIN,INPUT_PULLUP);
pinMode(REMOTE_RX_PIN,INPUT_PULLUP);
if(SAMPLE_RATE < 16000){
volMod = 3;
}else{
#if !defined (tenBit)
volMod = 2;
#endif
}
radio.setChannel(1); // Set RF channel to 1
radio.setAutoAck(0); // Disable ACKnowledgement packets
radio.setDataRate(RF_SPEED); // Set data rate as specified in user options
radio.setCRCLength(RF24_CRC_8);
radio.openWritingPipe(pipes[0]); // Set up reading and writing pipes. All of the radios write via multicast on the same pipe
radio.openReadingPipe(1,pipes[1]); // All of the radios listen by default to the same multicast pipe
radio.openReadingPipe(2,pipes[radioIdentifier + 2]); // Every radio also has its own private listening pipe
radio.startListening(); // NEED to start listening after opening a reading pipe
timerStart(); // Get the timer running
RX(); // Start listening for transmissions
#if !defined (MANUAL_BUTTON_HANDLING)
TIMSK0 |= _BV(OCIE0B);
#endif
}
//General functions for volume control
void vol(bool upDn){
if(upDn==1){ volMod++;}
else{ volMod--; }
}
void RF24Audio::volume(bool upDn){
vol(upDn);
}
void RF24Audio::setVolume(char vol) {
volMod = vol - 4 ;
}
void RF24Audio::timerStart(){
ICR1 = 10 * (1600000/SAMPLE_RATE); //Timer will count up to this value from 0;
TCCR1A = _BV(COM1A1) | _BV(COM1B0) | _BV(COM1B1); //Enable the timer port/pin as output
TCCR1A |= _BV(WGM11); //WGM11,12,13 all set to 1 = fast PWM/w ICR TOP
TCCR1B = _BV(WGM13) | _BV(WGM12) | _BV(CS10); //CS10 = no prescaling
}
#if !defined (MANUAL_BUTTON_HANDLING)
void handleButtons(){
boolean state = digitalRead(TX_PIN); //Get the state of the transmitting pin
if(!state){ //If the pin is low, start transmitting
#if !defined (RX_ONLY)
if(!transmitting){ //Only do this if not already transmitting
transmitting = 1; //Set the transmitting variable
TX(); //Switch to TX mode
}
}else{ //If the TX pin is low
#endif
if(transmitting){ //If still in transmitting mode
RX(); transmitting = 0; //Start receiving/Stop transmitting
}
}
if(!digitalRead(VOL_UP_PIN)){ //If the volume pin is high, raise the volume
if(millis()-volTime > 200){ //De-bounce of buttons by 200ms
vol(1); volTime = millis();
}
}else
if(!digitalRead(VOL_DN_PIN)){
if(millis()-volTime > 200){
vol(0); volTime = millis();
}
}
if(!digitalRead(REMOTE_TX_PIN)){ //Start remote recording and transmission of audio
if(!transmitting){ //If not transmitting already
radi.stopListening(); //Stop listening and write the request twice for good measure
radi.write(&txCmd,2);
radi.write(&txCmd,2);
radi.startListening();}
delay(200); //Delay as a simple button debounce
}else
if(!digitalRead(REMOTE_RX_PIN)){ //With TX timeout enabled, this will stop remote recording by stopping transmission for 2 seconds
delay(2000);
}
}
#endif //Manual button handling override
void rampDown(){
int current = OCR1A;
if(current > 0){
for(int i=0; i < ICR1; i++){
#if defined(rampMega)
OCR1B = constrain((current + i),0,ICR1);
OCR1A = constrain((current - i),0,ICR1);
#else
OCR1B = constrain((current - i),0,ICR1);
OCR1A = constrain((current - i),0,ICR1);
#endif
//for(int i=0; i<10; i++){ while(TCNT1 < ICR1-50){} }
delayMicroseconds(100);
}
}
}
void rampUp(byte nextVal){
/*int current = OCR1A;
if(current > 0){
for(int i=0; i < ICR1; i++){
#if defined(rampMega)
OCR1B = constrain((current - i),0,ICR1/2);
OCR1A = constrain((current + i),0,ICR1/2);
#else
OCR1B = constrain((current + i),0,ICR1/2);
OCR1A = constrain((current + i),0,ICR1/2);
#endif
//for(int i=0; i<10; i++){ while(TCNT1 < ICR1-50){} }
delayMicroseconds(100);
}
}*/
//digitalWrite(12,HIGH);
#if defined(rampMega)
unsigned int resolution = ICR1;
OCR1A = 0; OCR1B = resolution;
for(int i=0; i < resolution; ++i){
OCR1B = constrain(resolution-i,0,resolution);
//if(bitRead(*TCCRnB[tt],0)){
// for(int i=0; i<10; i++){
// while(*TCNT[tt] < resolution-50){}
// }
//}else{
//delayMicroseconds(10);
//__asm__("nop\n\t""nop\n\t""nop\n\t""nop\n\t");
//}
}
#endif
byte tmp = 200;
unsigned int mod;
if(volMod > 0){ mod = OCR1A >> volMod; }else{ mod = OCR1A << (volMod*-1); }
if(tmp > mod){
for(unsigned int i=0; i<buffSize; i++){ mod = constrain(mod+1,mod, tmp); buffer[0][i] = mod; }
for(unsigned int i=0; i<buffSize; i++){ mod = constrain(mod+1,mod, tmp); buffer[1][i] = mod; }
}else{
for(unsigned int i=0; i<buffSize; i++){ mod = constrain(mod-1,tmp ,mod); buffer[0][i] = mod; }
for(unsigned int i=0; i<buffSize; i++){ mod = constrain(mod-1,tmp, mod); buffer[1][i] = mod; }
}
whichBuff = 0; buffEmpty[0] = 0; buffEmpty[1] = 0; buffCount = 0;
}
void RF24Audio::transmit(){
TX();
}
void RF24Audio::receive(){
RX();
}
#if !defined MANUAL_BUTTON_HANDLING // Allows users to totally customize button handling or disable it
ISR(TIMER0_COMPB_vect){ // Non-blocking interrupt vector for button management. Is triggered ~1000 times/second by default on Arduino
handleButtons(); // Check for external button presses at the default timer0 rate
}
#endif
uint64_t RF24Audio::getAddress(byte addressNo){
return pipes[addressNo];
}
/*****************************************************************************************************************************/
/****************************************** Reception (RX) Section ***********************************************************/
void handleRadio(){
if(buffEmpty[!whichBuff] && streaming){ // If in RX mode and a buffer is empty, load it
if(radi.available() ){
boolean n=!whichBuff; // Grab the changing value of which buffer is not being read before enabling nested interrupts
TIMSK1 &= ~_BV(ICIE1); // Disable this interrupt so it is not triggered while still running (this may take a while)
sei(); // Enable nested interrupts (Other interrupts can interrupt this one)
radi.read(&buffer[n],32); // Read the payload from the radio
buffEmpty[n] = 0; // Indicate that a buffer is now full and ready to play
pauseCntr = 0; // For disabling speaker when no data is being received
TIMSK1 |= _BV(ICIE1); // Finished, re-enable the interrupt vector that runs this function
}else{ pauseCntr++; } // No payload available, keep track of how many for disabling the speaker
if(pauseCntr > 50){ // If we failed to get a payload 250 times, disable the speaker output
pauseCntr = 0; // Reset the failure counter
rampDown(); // Ramp down the speaker (prevention of popping sounds)
streaming = 0; // Indicate that streaming is stopped
TIMSK1 &= ~(_BV(TOIE1) ); // Disable the TIMER1 overflow vector (playback)
#if defined (ENABLE_LED)
TCCR0A &= ~_BV(COM0A1); // Disable the TIMER0 LED visualization
#endif
TCCR1A &= ~(_BV(COM1A1) | _BV(COM1B1) | _BV(COM1B0)); // Disable speaker output
}
}else
if(!streaming){ // If not actively reading a stream, read commands instead
if(radi.available() ){ // If a payload is available
TIMSK1 &= ~_BV(ICIE1); // Since this is called from an interrupt, disable it
sei(); // Enable global interrupts (nested interrupts) Other interrupts can interrupt this one
radi.read(&buffer[0],32); // Read the payload into the buffer
switch(buffer[0][0]){ // Additional commands can be added here for controlling other things via radio command
#if !defined (RX_ONLY)
case 'r': if(buffer[0][1] == 'R' && radioIdentifier < 2){ //Switch to TX mode if we received the remote tx command and this is radio 0 or 1
TX();
}
break;
#endif
default: streaming= 1; // If not a command, treat as audio data, enable streaming
TCCR1A |= _BV(COM1A1) | _BV(COM1B1) | _BV(COM1B0); //Enable output to speaker pin
rampUp(buffer[0][31]); // Ramp up the speakers to prevent popping
//buffEmpty[0] = false; // Set the status of the memory buffers
//buffEmpty[1] = true;
TIMSK1 |= _BV(TOIE1); // Enable the overflow vector
#if defined (ENABLE_LED)
TCCR0A |= _BV(COM0A1); // Enable the LED visualization output
#endif
break;
}
TIMSK1 |= _BV(ICIE1); // Finished: Re-enable the interrupt that runs this function.
}
}
}
void RX(){ // Start Receiving
TIMSK1 &= ~_BV(OCIE1B) | _BV(OCIE1A); // Disable the transmit interrupts
ADCSRA = 0; ADCSRB = 0; // Disable Analog to Digital Converter (ADC)
buffEmpty[0] = 1; buffEmpty[1] = 1; // Set the buffers to empty
#if defined (oversampling) // If running the timer at double the sample rate
ICR1 = 10 * (800000/SAMPLE_RATE); // Set timer top for 2X oversampling
#else
ICR1 = 10 * (1600000/SAMPLE_RATE); // Timer running at normal sample rate speed
#endif
radi.openWritingPipe(pipes[0]); // Set up reading and writing pipes
radi.openReadingPipe(1,pipes[1]);
radi.startListening(); // Exit sending mode
TIMSK1 = _BV(ICIE1); // Enable the capture interrupt vector (handles buffering and starting of playback)
}
boolean nn = 0;
volatile byte bufCtr = 0;
volatile unsigned int visCtr = 0;
ISR(TIMER1_CAPT_vect){ // This interrupt checks for data at 1/16th the sample rate. Since there are 32 bytes per payload, it gets two chances for every payload
bufCtr++; visCtr++; // Keep track of how many times the interrupt has been triggered
if(bufCtr >= 16){ // Every 16 times, do this
handleRadio(); // Check for incoming radio data if not transmitting
bufCtr = 0; // Reset this counter
if(visCtr >= 32 && streaming){ // Run the visualisation at a much slower speed so we can see the changes better
OCR0A = buffer[whichBuff][0] << 2; // Adjust the TIMER0 compare match to change the PWM duty and thus the brightess of the LED
visCtr = 0; // Reset the visualization counter
}
}
}
// Receiving interrupt
ISR(TIMER1_OVF_vect){ // This interrupt vector loads received audio samples into the timer register
if(buffEmpty[whichBuff] ){ whichBuff=!whichBuff; }else{ // Return if both buffers are empty
#if defined (oversampling) // If running the timers at 2X speed, only load a new sample every 2nd time
if(lCntr){lCntr = !lCntr;return;} lCntr=!lCntr;
#endif
#if !defined (tenBit)
/*************** Standard 8-Bit Audio Playback ********************/
if(volMod < 0 ){ //Load an audio sample into the timer compare register
OCR1A = OCR1B = (buffer[whichBuff][intCount] >> volMod*-1); //Output to speaker at a set volume
}else{
OCR1A = OCR1B = buffer[whichBuff][intCount] << volMod;
}
intCount++; //Keep track of how many samples have been loaded
if(intCount >= buffSize){ //If the buffer is empty, do the following
intCount = 0; //Reset the sample count
buffEmpty[whichBuff] = true; //Indicate which buffer is empty
whichBuff = !whichBuff; //Switch buffers to read from
}
/*************** 10 - bit Audio *************************************/
#else
sampl = buffer[whichBuff][intCount]; // Load 8bits of the sample into a temporary buffer
bitWrite( sampl, 8, bitRead( buffer[whichBuff][bytePos],bitPos)); // Read the 9th bit, and write it to the temporary buffer
bitPos++; // Keep track of the bit-position, since bits are loaded in bytes 25-31
bitWrite(sampl, 9, bitRead(buffer[whichBuff][bytePos],bitPos)); // Read the 10th bit, and write it to the temporary buffer
bitPos++; // Keep track of the bit-position
if(volMod < 0 ){ // Load an audio sample into the timer compare register
OCR1A = OCR1B = sampl >> (volMod*-1); // Output to speaker at a set volume
}else{
OCR1A = OCR1B = sampl << volMod;
}
if(bitPos >=8){ // If 8 bits have been read, reset the counter and switch to the next byte
bitPos = 0;
bytePos = bytePos+1;
}
intCount++; // Keep track of how many samples have been loaded
if(intCount >= 25){ // If the buffer is empty, do the following
bytePos = 25;
bitPos = 0;
intCount = 0; // Reset the sample count
buffEmpty[whichBuff] = true; // Indicate which buffer is empty
whichBuff = !whichBuff; // Switch buffers to read from
}
#endif
}
}
/*****************************************************************************************************************************/
/*************************************** Transmission (TX) Section ***********************************************************/
#if !defined (RX_ONLY) //If TX is enabled:
void RF24Audio::broadcast(byte radioID){
if(radioID == radioIdentifier){ return; } // If trying to send to our own address, return
noInterrupts(); // Disable interrupts during change of transmission address
if(radioID == broadcastVal){
radio.openWritingPipe(pipes[1]); // Use the public multicast pipe
}else{
radio.openWritingPipe(pipes[radioID + 2]); // Open a pipe to the specified radio(s). If two or more radios share the same ID,
} // they will receive the same single broadcasts, but will not be able to initiate
interrupts(); // private communication betwen each other
}
//Transmission sending interrupt
ISR(TIMER1_COMPA_vect){ // This interrupt vector sends the samples when a buffer is filled
if(buffEmpty[!whichBuff] == 0){ // If a buffer is ready to be sent
a = !whichBuff; // Get the buffer # before allowing nested interrupts
TIMSK1 &= ~(_BV(OCIE1A)); // Disable this interrupt vector
sei(); // Allow other interrupts to interrupt this vector (nested interrupts)
//radi.startFastWrite(&buffer[a],32);
radi.writeFast(&buffer[a],32);
buffEmpty[a] = 1; // Mark the buffer as empty
TIMSK1 |= _BV(OCIE1A); // Re-enable this interrupt vector
}
}
//Transmission buffering interrupt
ISR(TIMER1_COMPB_vect){ // This interrupt vector captures the ADC values and stores them in a buffer
#if !defined (tenBit) // 8-bit samples
buffer[whichBuff][buffCount] = bytH = ADCH; // Read the high byte of the ADC register into the buffer for 8-bit samples
#if defined (speakerTX) // If output to speaker while transmitting is enabled
if(volMod < 0 ){ OCR1A = bytH >> (volMod*-1); // Output to speaker at a set volume if defined
}else{ OCR1A = bytH << volMod;
}
#endif
#else // 10-bit samples are a little more complicated, but offer better quality for lower sample rates
buffer[whichBuff][buffCount] = bytL = ADCL; // In 10-bit mode, the ADC register is right-adjusted, we need to read the low, then high byte each time
bytH = ADCH;
bitWrite(buffer[whichBuff][bytePos],bitPos, bitRead(bytH,0)); // The low bytes are stored in the first 25 bytes of the payload. The additional 2 bits are stored in
bitWrite(buffer[whichBuff][bytePos],bitPos+1, bitRead(bytH,1));// pairs in the remaining bytes #25 to 31. Read the first and 2nd bits of the high register into the payload.
bitPos+=2;
if(bitPos >= 8){ bitPos = 0; bytePos = bytePos+1; } // Every time a byte is full, increase byte position by 1 and reset the bit count.
#if defined (speakerTX) // If output to speaker while transmitting is enabled
sampl = bytL; // Load the two bytes into the 2-byte unsigned integer. Low byte first
sampl |= bytH << 8; // Shift the high byte 8bits and load it into the unsigned int using an OR comparison
if(volMod < 0 ){ OCR1A = sampl >> (volMod*-1); // Output to speaker at a set volume if defined
}else{ OCR1A = sampl << volMod;
}
#endif
#endif
buffCount++; // Keep track of how many samples have been loaded
#if !defined (tenBit) // 8-bit mode
if(buffCount >= 32){ // In 8-bit mode, do this every 32 samples
#else // 10-bit mode
if(buffCount >= 25){ // In 10-bit mode, do this every 25 samples
bytePos = 25; // Reset the position for the extra 2 bits to the 25th byte
bitPos = 0; // Reset the bit position for the extra 2 bits
#endif
//Both modes
buffCount = 0; // Reset the sample counter
buffEmpty[!whichBuff] = 0; // Indicate which bufffer is ready to send
whichBuff = !whichBuff; // Switch buffers and load the other one
}
}
void TX(){
TIMSK1 &= ~(_BV(ICIE1) | _BV(TOIE1)); // Disable the receive interrupts
#if defined (ENABLE_LED)
TCCR0A &= ~_BV(COM0A1); // Disable LED visualization
#endif
radi.openWritingPipe(pipes[1]); // Set up reading and writing pipes
radi.openReadingPipe(1,pipes[0]);
radi.stopListening(); // Enter transmit mode on the radio
streaming = 0;
buffCount = 0; buffEmpty[0] = 1; buffEmpty[1] = 1; //Set some variables
byte pin = ANALOG_PIN;
/*** This section taken from wiring_analog.c to translate between pins and channel numbers ***/
#if defined(analogPinToChannel)
#if defined(__AVR_ATmega32U4__)
if (pin >= 18) pin -= 18; // allow for channel or pin numbers
#endif
pin = analogPinToChannel(pin);
#elif defined(__AVR_ATmega1280__) || defined(__AVR_ATmega2560__)
if (pin >= 54) pin -= 54; // allow for channel or pin numbers
#elif defined(__AVR_ATmega32U4__)
if (pin >= 18) pin -= 18; // allow for channel or pin numbers
#elif defined(__AVR_ATmega1284__) || defined(__AVR_ATmega1284P__) || defined(__AVR_ATmega644__) || defined(__AVR_ATmega644A__) || defined(__AVR_ATmega644P__) || defined(__AVR_ATmega644PA__)
if (pin >= 24) pin -= 24; // allow for channel or pin numbers
#else
if (pin >= 14) pin -= 14; // allow for channel or pin numbers
#endif
#if defined(ADCSRB) && defined(MUX5)
// the MUX5 bit of ADCSRB selects whether we're reading from channels
// 0 to 7 (MUX5 low) or 8 to 15 (MUX5 high).
ADCSRB = (ADCSRB & ~(1 << MUX5)) | (((pin >> 3) & 0x01) << MUX5);
#endif
#if defined(ADMUX)
ADMUX = (pin & 0x07) | _BV(REFS0); //Enable the ADC PIN and set 5v Analog Reference
#endif
ICR1 = 10 * (1600000/SAMPLE_RATE); //Timer counts from 0 to this value
#if !defined (speakerTX) //If disabling/enabling the speaker, ramp it down
rampDown();
TCCR1A &= ~(_BV(COM1A1)); //Disable output to speaker
#endif
TCCR1A &= ~(_BV(COM1B1) |_BV(COM1B0) );
#if !defined (tenBit)
ADMUX |= _BV(ADLAR); //Left-shift result so only high byte needs to be read
#else
ADMUX &= ~_BV(ADLAR); //Don't left-shift result in 10-bit mode
#endif
ADCSRB |= _BV(ADTS0) | _BV(ADTS0) | _BV(ADTS2); //Attach ADC start to TIMER1 Capture interrupt flag
byte prescaleByte = 0;
if( SAMPLE_RATE < 8900){ prescaleByte = B00000111;} //128
else if( SAMPLE_RATE < 18000){ prescaleByte = B00000110;} //ADC division factor 64 (16MHz / 64 / 13clock cycles = 19230 Hz Max Sample Rate )
else if( SAMPLE_RATE < 27000){ prescaleByte = B00000101;} //32 (38461Hz Max)
else if( SAMPLE_RATE < 65000){ prescaleByte = B00000100;} //16 (76923Hz Max)
else { prescaleByte = B00000011;} //8 (fast as hell)
ADCSRA = prescaleByte; //Adjust sampling rate of ADC depending on sample rate
ADCSRA |= _BV(ADEN) | _BV(ADATE); //ADC Enable, Auto-trigger enable
TIMSK1 = _BV(OCIE1B) | _BV(OCIE1A); //Enable the TIMER1 COMPA and COMPB interrupts
}
#endif
|
Mental health nurses' views on therapeutic optimism.
Registered nurses (RN) coordinate acute mental health units on a 24-hour basis and it behoves researchers to actually ask these nurses what they think contributes to their ability to work with patients in optimistic ways. In this study, 40 RN working in acute mental health settings were asked a series of questions to explore positive aspects of nursing work, which includes therapeutic optimism. Three themes were identified: (i) different ways nurses foster therapeutic optimism; (ii) perceptions of how an optimistic environment is fostered, and (iii) improvement of ward culture. Findings show the pivotal role mental health nurses have in improving teamwork, good communication, sharing, and collaboration, in addition to preceptoring and supervision. Furthermore, effective clinical management is essential to therapeutic optimism and, in this research, is considered to be the aspect of acute mental health nursing most relevant to improving the ward culture.
|
---
author:
- |
Hwancheol Jeong, , Jongjeong Kim, Weonjong Lee, Jeonghwan Pak, Sungwoo Park\
Lattice Gauge Theory Research Center, CTP, and FPRD,\
Department of Physics and Astronomy,\
Seoul National University, Seoul, 151-747, South Korea\
E-mail:
- SWME Collaboration
bibliography:
- 'ref.bib'
title: 'Non-perturbative Renormalization of Four-Fermion Operators Relevant to with Staggered Quarks'
---
Introduction
============
The indirect CP violation parameter, in the neutral kaon system is very well known with $\approx 0.5\,\%$ precision from experiments [@Agashe:2014kda]. Our theoretical estimate of directly from the standard model (SM) has 3.4$\sigma$ tension with the experiment in the exclusive $V_{cb}$ channel [@Lee:2014bjl]. In order to widen the gap in the unit of $\sigma$, we need to increase the precision of the calculation of $B_K$ and $V_{cb}$ in lattice QCD. In our calculation, one of the dominant source of error comes from the matching factor for ($\approx 4.4\%$) using the one-loop perturbation theory. Hence, it becomes essential to reduce the matching factor error. The non-perturbative renormalization method (NPR) with the RI-MOM [@Aoki:2007xm] can reduce this error down to the $\approx 2\%$ level. In the previous work of Refs. [@Kim:2013bta; @Lytle:2013qoa], the NPR method has been applied to the staggered bilinear operators. Here, we present preliminary results of the renormalization factors of four-fermion operators relevant to operator obtained using NPR in the RI-MOM scheme with improved staggered fermions.
Four-fermion operator renormalization in the RI-MOM scheme
==========================================================
There are two kinds of color contraction of four-fermion operators. A general one-color trace four-fermion operator is defined as follows. $$\begin{aligned}
O_{\alpha, \I}(y)
&= \sum_{\substack{A,B,\\C,D}}\sum_{\substack{c_{1},c_{2},\\c_{3},c_{4}}}
\big[
\ol{\chi}_{c_1}(y_A) \ol{(\gamma_{S_1} \otimes \xi_{F_1})}_{AB} \chi_{c_2}(y_B)
\big]
\big[
\ol{\chi}_{c_3}(y_C) \ol{(\gamma_{S_2} \otimes \xi_{F_2})}_{CD} \chi_{c_4}(y_D)
\big]
\nonumber \\
& \qquad U_{AD;c_{1}c_{4}}(y)U_{BC;c_{2}c_{3}}(y) \,,\end{aligned}$$ and a general two-color trace four-fermion operator is defined as follows. $$\begin{aligned}
O_{\alpha, \II}(y)
&= \sum_{\substack{A,B,\\C,D}}\sum_{\substack{c_{1},c_{2},\\c_{3},c_{4}}}
\big[
\ol{\chi}_{c_1}(y_A) \ol{(\gamma_{S_1} \otimes \xi_{F_1})}_{AB} \chi_{c_2}(y_B)
\big]
\big[
\ol{\chi}_{c_3}(y_C) \ol{(\gamma_{S_2} \otimes \xi_{F_2})}_{CD} \chi_{c_4}(y_D)
\big]
\nonumber \\
& \qquad U_{AB;c_{1}c_{2}}(y)U_{CD;c_{3}c_{4}}(y)\,,\end{aligned}$$ where $\alpha = [S_1 \otimes F_1][S_2 \otimes F_2]$ is an operator index, and $c_i$ are color indices. The $y$ represents a coordinate of the hypercube with its lattice spacing $2a$. The indices $A$, $B$, $C$ and $D$ are hypercubic vectors: for example, $A = (1,1,0,0)$. Here, we use the notation of $y_A = 2y + A$. $U_{AB;c_1 c_2}(y)$ is a gauge link, an average of the shortest paths which connect $y_A$ and $y_B$ as products of HYP-smeared fat links. $\gamma_{S}$ represents the spin and $\xi_{F}$ the taste. Here, $\chi(y_B)$ represents HYP-smeared staggered quark field. We calculate the amputated Green’s function using same method introduced in Ref. [@Kim:2013bta].
As an example, we choose the four fermion operators used to calculate $B_K$ in order to illustrate how the NPR method produces the matching factors. We introduce the following simple notations for the operators. $$\begin{aligned}
& O_{1} \equiv O_{[V \otimes P][V \otimes P], \I} \,, \qquad
O_{2} \equiv O_{[V \otimes P][V \otimes P], \II}\,,
\nonumber \\
& O_{3} \equiv O_{[A \otimes P][A \otimes P], \I} \,, \qquad
O_{4} \equiv O_{[A \otimes P][A \otimes P], \II} \,.
\label{eq:op}\end{aligned}$$ First, we divide the lattice operators into two classes: (C) the diagonal operators defined in Eq., {$O_1$, $O_2$, $O_3$, $O_4$}, which have the $\xi_{5}$ tastes in both bilinears, and (D) the off-diagonal operators which are remaining operators with taste different from $\xi_{5}$. The tree level operator is sum of the operators in the (C) class. $$\begin{aligned}
O^{\tree}_{B_K} =
O^{\tree}_{1} +
O^{\tree}_{2} +
O^{\tree}_{3} +
O^{\tree}_{4}\end{aligned}$$ The projection operators are also defined in the same way in Eq. as follows. $$\begin{aligned}
\mathbb{P}_{1} &\equiv
\frac{1}{N}
\ol{\ol{(\gamma_{\mu}^{\dagger} \otimes \xi_{5}^{\dagger})}}_{BA}
\ol{\ol{(\gamma_{\mu}^{\dagger} \otimes \xi_{5}^{\dagger})}}_{DC}
\delta_{c_4 c_1}\delta_{c_3 c_2} \,,&
%
\mathbb{P}_{2} &\equiv
\frac{1}{N}
\ol{\ol{(\gamma_{\mu}^{\dagger} \otimes \xi_{5}^{\dagger})}}_{BA}
\ol{\ol{(\gamma_{\mu}^{\dagger} \otimes \xi_{5}^{\dagger})}}_{DC}
\delta_{c_2 c_1}\delta_{c_4 c_3} \,,
%
\nonumber\\
%
\mathbb{P}_{3} &\equiv
\frac{1}{N}
\ol{\ol{(\gamma_{\mu 5}^{\dagger} \otimes \xi_{5}^{\dagger})}}_{BA}
\ol{\ol{(\gamma_{\mu 5}^{\dagger} \otimes \xi_{5}^{\dagger})}}_{DC}
\delta_{c_4 c_1}\delta_{c_3 c_2} \,,&
%
\mathbb{P}_{4} &\equiv
\frac{1}{N}
\ol{\ol{(\gamma_{\mu 5}^{\dagger} \otimes \xi_{5}^{\dagger})}}_{BA}
\ol{\ol{(\gamma_{\mu 5}^{\dagger} \otimes \xi_{5}^{\dagger})}}_{DC}
\delta_{c_2 c_1}\delta_{c_4 c_3} \,.\end{aligned}$$ Here, $N$ is normalization factor given as follows. $$\begin{aligned}
N = 3072 =
\underbrace{4^4}_{\text{spin}}
\times
\underbrace{4^4}_{\text{taste}}
\times
(
\underbrace{3}_{\text{1-color trace}} +
\underbrace{9}_{\text{2-color trace}}
)\end{aligned}$$ We fix the normalization factor $N$ such that, when we apply the projection operators to the tree level amputated Green’s function, it satisfies the following conditions. $$\begin{aligned}
\label{eq:tree:1}
&\tr[\Lambda^{\tree}_{B_K} \mathbb{P}_{1}]
%
= \tr[\Lambda^{\tree}_{B_K} \mathbb{P}_{2}]
%
=\tr[\Lambda^{\tree}_{B_K} \mathbb{P}_{3}]
%
=\tr[\Lambda^{\tree}_{B_K} \mathbb{P}_{4}]
%
=1 \\
%
&\tr[\Lambda^{\tree}_{B_K} \mathbb{P}_{(D)}]
%
=\tr[\Lambda^{\tree}_{(D)} \mathbb{P}_{(C)}]
%
=0
\label{eq:tree:2}\end{aligned}$$ Here, note that the diagonal terms equal to one and the off-diagonal terms becomes zero.
The renormalized operator is defined as follows. $$\begin{aligned}
O^{R}_{B_K}
= z_{1}O^{B}_{1} + z_{2}O^{B}_{2}
+ z_{3}O^{B}_{3} + z_{4}O^{B}_{4}
+ \sum_{\alpha \in (D)} z_{\alpha}O^{B}_{\alpha}\,,\end{aligned}$$ where the superscript R (B) denotes renormalized (bare) quantity, and the coefficients $z_i$ are renormalization factors. The renormalization of quark fields is defined as follows. $$\begin{aligned}
\chi^{R} = Z^{1/2}_{q} \chi^{B}\end{aligned}$$ The amputated Green’s function is obtained by multiplying the inverse propagators to the unamputated Green’s function. Hence the renormalized amputated Green’s function is as follows. $$\begin{aligned}
\label{eq:renorm}
\Lambda^{R}_{B_K}
= \frac{z_1}{z_q^2}\Lambda^{B}_{1}
+ \frac{z_2}{z_q^2}\Lambda^{B}_{2}
+ \frac{z_3}{z_q^2}\Lambda^{B}_{3}
+ \frac{z_4}{z_q^2}\Lambda^{B}_{4}
+ \sum_{\alpha \in (D)} \frac{z_{\alpha}}{z^2_q}\Lambda^{B}_{\alpha}\end{aligned}$$ The RI-MOM scheme prescription is that the renormalized quantity is equal to its tree level value. $$\begin{aligned}
\label{eq:RI-MOM}
\tr[\Lambda^{R}_\alpha(\wtd{p},\wtd{p},\wtd{p},\wtd{p})
\mathbb{P}_\beta]
& = \tr[\Lambda^{\tree}_\alpha(\wtd{p},\wtd{p},\wtd{p},\wtd{p})
\mathbb{P}_\beta] \,,\end{aligned}$$ where $\wtd{p}$ is a momentum defined in the reduced Brillouin zone.[^1] We define the projected amputated Green’s function as follows. $$\begin{aligned}
\Gamma^{B}_{\alpha \beta} &\equiv
\frac{1}{z^2_q}\tr[\Lambda^{B}_{\alpha}\mathbb{P}_{\beta}]\end{aligned}$$ Hence, from Eq. , Eq. , Eq. and Eq. , we obtain the following relations. $$\begin{aligned}
1 &=
z_{1}\Gamma^{B}_{1\alpha} +
z_{2}\Gamma^{B}_{2\alpha} +
z_{3}\Gamma^{B}_{3\alpha} +
z_{4}\Gamma^{B}_{4\alpha} +
\sum_{\gamma \in (D)}
z_{\gamma}\Gamma^{B}_{\gamma \alpha}
\,, \quad
\alpha \in (C) \\
%
0 &=
z_{1}\Gamma^{B}_{1\beta} +
z_{2}\Gamma^{B}_{2\beta} +
z_{3}\Gamma^{B}_{3\beta} +
z_{4}\Gamma^{B}_{4\beta} +
\sum_{\gamma \in (D)}
z_{\gamma}\Gamma^{B}_{\gamma \beta}
\,, \quad
\beta \in (D)\end{aligned}$$ We can express these equations as a matrix equation. $$\begin{aligned}
\vec{z}_{\tree} = \vec{z} \cdot \hat{\Gamma}^{B} \,, \end{aligned}$$ where $\vec{z}_{\tree}$ and $\vec{z}$ are vectors as follows. $$\begin{aligned}
&\vec{z}_{\tree} = (1, 1, 1, 1, 0, \cdots, 0)\,, \quad
\vec{z} = (z_{1}, z_{2}, z_{3}, z_{4}, z_5, z_6, \cdots)\end{aligned}$$ where $z_i$ with $i\geq 5$ are renormalization factors of off-diagonal operators. The $\hat{\Gamma}^{B}$ is a matrix as follows. The upper-left (red) block elements are diagonal terms and others are off-diagonal terms. $$\hat{\Gamma}^{B} = \left(
\begin{array}{c c c c :c c c}
\textcolor{red}{\Gamma^{B}_{11}} &
\textcolor{red}{\Gamma^{B}_{12}} &
\textcolor{red}{\Gamma^{B}_{13}} &
\textcolor{red}{\Gamma^{B}_{14}} &
\Gamma^{B}_{15} &
\Gamma^{B}_{16} &
\cdots \\
%
\textcolor{red}{\Gamma^{B}_{21}} &
\textcolor{red}{\Gamma^{B}_{22}} &
\textcolor{red}{\Gamma^{B}_{23}} &
\textcolor{red}{\Gamma^{B}_{24}} &
\Gamma^{B}_{25} &
\Gamma^{B}_{26} &
\cdots \\
%
\textcolor{red}{\Gamma^{B}_{31}} &
\textcolor{red}{\Gamma^{B}_{32}} &
\textcolor{red}{\Gamma^{B}_{33}} &
\textcolor{red}{\Gamma^{B}_{34}} &
\Gamma^{B}_{35} &
\Gamma^{B}_{36} &
\cdots \\
%
\textcolor{red}{\Gamma^{B}_{41}} &
\textcolor{red}{\Gamma^{B}_{42}} &
\textcolor{red}{\Gamma^{B}_{43}} &
\textcolor{red}{\Gamma^{B}_{44}} &
\Gamma^{B}_{45} &
\Gamma^{B}_{46} &
\cdots \\
%
\hdashline
%
\Gamma^{B}_{51} &
\Gamma^{B}_{52} &
\Gamma^{B}_{53} &
\Gamma^{B}_{54} &
\Gamma^{B}_{55} &
\Gamma^{B}_{56} &
\cdots \\
%
\Gamma^{B}_{61} &
\Gamma^{B}_{62} &
\Gamma^{B}_{63} &
\Gamma^{B}_{64} &
\Gamma^{B}_{65} &
\Gamma^{B}_{66} &
\cdots \\
%
\vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \ddots
\end{array}
\right)$$ Hence, we can compute $z$-factors from the inverse of $\hat{\Gamma}^{B}$ matrix as follows. $$\begin{aligned}
\vec{z} = \vec{z}_{\tree} \cdot (\hat{\Gamma}^{B})^{-1}\end{aligned}$$ The $\hat{\Gamma}^{B}$ can be rewritten by sub-matrices as follows. $$\begin{aligned}
\hat{\Gamma}^B =
\begin{pmatrix}
X_{4 \times 4} & Y_{4 \times 20} \\
Z_{20 \times 4} & W_{20 \times 20}
\end{pmatrix}\end{aligned}$$ Here, $X$ is diagonal terms, and $Y$ is off-diagonal terms. $$\begin{aligned}
X =
\begin{pmatrix}
\Gamma^{B}_{11} & \Gamma^{B}_{12} & \Gamma^{B}_{13} & \Gamma^{B}_{14} \\
\Gamma^{B}_{21} & \Gamma^{B}_{22} & \Gamma^{B}_{23} & \Gamma^{B}_{24} \\
\Gamma^{B}_{31} & \Gamma^{B}_{32} & \Gamma^{B}_{33} & \Gamma^{B}_{34} \\
\Gamma^{B}_{41} & \Gamma^{B}_{42} & \Gamma^{B}_{43} & \Gamma^{B}_{44}
\end{pmatrix} \,,
\qquad
Y =
\begin{pmatrix}
\Gamma^{B}_{15} & \Gamma^{B}_{16} & \cdots \\
\Gamma^{B}_{25} & \Gamma^{B}_{26} & \cdots \\
\Gamma^{B}_{35} & \Gamma^{B}_{36} & \cdots \\
\Gamma^{B}_{45} & \Gamma^{B}_{46} & \cdots
\end{pmatrix}\end{aligned}$$ The number of the off-diagonal operators is 20 and they are $\{O_5, O_6, \cdots\} =
\{
(S \otimes V)(S \otimes V)_{\I},
(S \otimes V)(S \otimes V)_{\II},
(S \otimes A)(S \otimes A)_{\I},
\cdots\}$. We assume that $Z \simeq Y^{T} \simeq
\mathcal{O}(\alpha_s)$[^2] and $W \simeq 1 + \mathcal{O}(\alpha_s)$. The inverse of block matrix is $$\begin{aligned}
&(\hat{\Gamma}^B)^{-1} =
%
\begin{pmatrix}
(X - Y W^{-1} Z)^{-1}
%
& - X^{-1} Y (W - Z X^{-1} Y)^{-1} \\
%
- W^{-1} Z (X - Y W^{-1} Z)^{-1}
%
& (W - Z X^{-1} Y)^{-1}
\end{pmatrix} \,.\end{aligned}$$ Using power series expansion in $Y$ and $Z$, it becomes as follows. $$\begin{aligned}
%
(\hat{\Gamma}^B)^{-1} &\simeq
\begin{pmatrix}
%
X^{-1} + X^{-1} Y W^{-1} Z X^{-1} &
- X^{-1} Y (W^{-1} + W^{-1} Z X^{-1} Y W^{-1}) \\
%
- W^{-1} Z (X^{-1} + X^{-1} Y W^{-1} Z X^{-1}) &
W^{-1} + W^{-1} Z X^{-1} Y W^{-1}
%
\end{pmatrix}\end{aligned}$$ With our assumption, $Z \simeq Y^{T}$ and $W \simeq 1$, $$\begin{aligned}
(\hat{\Gamma}^B)^{-1} &\simeq
\begin{pmatrix}
%
X^{-1} + X^{-1} Y Y^{T} X^{-1} &
- X^{-1} Y (1 + Y^T X^{-1} Y ) \\
%
- Y^T (X^{-1} + X^{-1} Y Y^T X^{-1}) &
1 + Y^T X^{-1} Y
%
\end{pmatrix} \,.\end{aligned}$$
Results
=======
First, we present the data analysis of diagonal terms. Let us consider the element $\Gamma^{B}_{11}$ of $\hat{\Gamma}^B$ matrix. We measure the data for 5 valence quark masses and 9 external momenta. The scale of raw data is determined by external momentum ($\mu =
|\wtd{p}|$). Hence, we convert the scale of raw data to the common scale $\mu_{0}=
3\GeV$ using two-loop RG evolution [@Kim:2014tda]. We fit the data with respect to quark mass for a fixed momentum to the following function $f$ suggested from Ref. [@Blum:2001sr] based on the Weinberg theorem [@Weinberg:1959nj]. $$\begin{aligned}
f(m, a, \wtd{p})
= c_{1}
+ c_{2} \cdot (am)
+ c_{3} \cdot \frac{1}{(am)}
+ c_{4} \cdot \frac{1}{(am)^2},\end{aligned}$$ where $m$ is a valence quark mass. After m-fit, we take the $c_1(a\wtd{p})$ as chiral limit values. Because of the sea quark determinant contributions ($c_{3}, c_{4}
\propto (m^2_{\ell}m_s)^\nu$) with $\nu$ the number of zero modes, these pole terms contributions vanish in the chiral limit. The fitting results are presented in Table \[tab:m-fit\] and Fig. \[fig:m-fit\].
$\mu_0$ $c_1$ $c_2$ $c_3$ $c_4$ $\chi^2/\text{dof}$
--------- ------------- ------------- --------------- ------------------ ---------------------
3GeV 0.17991(20) -0.0975(15) 0.0007118(85) -0.000000790(36) 0.00194(40)
: \[tab:m-fit\] m-fit
We fit $c_1(a\wtd{p})$ to the following fitting function. $$\begin{aligned}
g(a\wtd{p}) &=
b_{1} +
b_{2} \cdot (a\wtd{p})^2 +
b_{3} \cdot ((a\wtd{p})^2)^2 +
b_{4} \cdot (a\wtd{p})^4 +
b_{5} \cdot ((a\wtd{p})^2)^3\end{aligned}$$ To avoid non-perturbative effects at small $(a \wtd{p} )^2 \leq 1$, we choose the momentum window as $(a \wtd{p} )^2 > 1$. Because we assume that those terms of $\mathcal{O}((a\wtd{p})^2)$ and higher order are pure lattice artifacts, we take the $b_1$ as $\Gamma^{B}_{11}$ value at $\mu_0=3\GeV$ in the RI-MOM scheme. The fitting result and plot are presented in Table \[tab:p-fit\] and Fig. \[fig:p-fit\].
$\mu_0$ $b_1$ $b_2$ $b_3$ $b_4$ $b_{5}$ $\chi^2/\text{dof}$
--------- ----------- ------------ ------------ ------------ -------------- ---------------------
3GeV 1.088(16) -0.515(18) 0.0953(74) 0.0020(65) -0.00663(91) 0.08(17)
: \[tab:p-fit\] p-fit
Similarly, we analyse the whole elements of $\hat{\Gamma}^{B}$ matrix. Results of diagonal terms in the inverse of $\hat{\Gamma}^{B}$ are $$\begin{aligned}
%
X^{-1}&=
\begin{pmatrix}
1.333(32) & -0.793(45) & 0.336(21) & 0.007(31) \\
-0.726(44) & 1.940(41) & -0.008(30) & -0.047(33) \\
0.341(30) & 0.032(45) & 1.222(32) & -0.604(37) \\
0.018(45) & -0.084(52) & -0.637(39) & 1.543(36)
\end{pmatrix} \\
X^{-1} Y Y^{T} X^{-1}
&=\begin{pmatrix}
0.0127(15) & -0.0079(11) & 0.0020(17) & -0.0001(10) \\
-0.00713(92) & 0.0064(11) & -0.0010(11) & -0.00045(73) \\
0.0020(18) & 0.0000(15) & 0.0200(90) & -0.0103(48) \\
0.0001(12) & -0.0012(11) & -0.0109(52) & 0.0059(28)
\end{pmatrix}\end{aligned}$$ We obtain $\vec{z}$ in RI-MOM scheme at $\mu_0=3\GeV$. We convert the scheme from RI-MOM to $\MSb$ using two-loop RG evolution. Results are summarized in Table \[tab:rg-evol\].
RI-MOM(3GeV) $\MSb$(3GeV)
------- -------------- --------------
$z_1$ 0.9666(78) 0.9812(79)
$z_2$ 1.095(30) 1.111(31)
$z_3$ 0.9139(73) 0.9277(74)
$z_4$ 0.898(28) 0.912(29)
: The $z$-factors in RI-MOM and $\MSb$ scheme. Here, the errors are purely statistical.[]{data-label="tab:rg-evol"}
Now let us switch the gear to the systematic errors. The first systematic error comes from the diagonal correction terms, $X^{-1} Y W^{-1} Z X^{-1} \approx X^{-1} Y Y^{T} X^{-1}$. We quote $E_{diag} \equiv \vec{z}_\tree \cdot X^{-1} Y Y^{T} X^{-1}$ as this error. The second systematic error comes from the off-diagonal correction terms. Their size ($- X^{-1} Y$) are typically less than $7\%$. However, thanks to the wrong taste suppression ($\ll 1\%$) [@Wlee:thesis], their effect becomes $\ll 0.07\%$. Hence we neglect them without loss of generality. Another systematic error comes from truncated higher order of the two-loop RG evolution factor (RI-MOM$\to\MSb$). We quote $E_{t} \equiv z_{i} \cdot \alpha_s^3$ as this error.
$\MSb$(3GeV) $E_{diag}$ $E_{t}$ $E_{tot}$
------- -------------- ------------ --------- -----------
$z_1$ 0.9812(79) 0.0077 0.0144 0.0163
$z_2$ 1.111(31) 0.0027 0.0163 0.0165
$z_3$ 0.9277(74) 0.0101 0.0136 0.0171
$z_4$ 0.912(29) 0.0050 0.0134 0.0143
: The systematic errors of $z$-factors in $\MSb$ scheme at $3\GeV$. $E_{tot}$ represents the total systematic error.[]{data-label="tab:sys-err"}
We add these systematic errors ($E_{diag}$ and $E_t$) in quadrature as summarized in Table \[tab:sys-err\].
In addition, we compare the NPR result($\MSb$ \[NDR\]) with those of one-loop perturbative matching. We quote truncated two-loop uncertainty: $E_{t}^{\text{one-loop}}
\equiv z_{i} \cdot \alpha_{s}^2$ as our estimate of the systematic error of one-loop matching.
NPR(3GeV) one-loop(3GeV) $\Delta$
------- --------------- ---------------- ---------------
$z_1$ 0.981(8)(16) 1.035(62) 0.83 $\sigma$
$z_2$ 1.111(31)(17) 1.120(67) 0.12 $\sigma$
$z_3$ 0.928(7)(17) 1.043(63) 1.75 $\sigma$
$z_4$ 0.912(29)(14) 0.953(57) 0.63 $\sigma$
: Comparison NPR results and one-loop perturbative matching factors at $3\GeV$.
Here, note that the results of NPR are consistent with those of one-loop matching within $2\sigma$. This indicates that our NPR results are quite reasonable.
Acknowledgments
===============
The research of W. Lee is supported by the Creative Research Initiatives Program (No. 2014001852) of the NRF grant funded by the Korean government (MEST). W. Lee would like to acknowledge the support from KISTI supercomputing center through the strategic support program for the supercomputing application research (KSC-2013-G2-005).
[^1]: Please refer to Ref. [@Kim:2013bta] for more details.
[^2]: Note that this approximation is good within factor of 3.
|
/35
Evaluate (-470)/(-25) + (-3198)/(-2665).
20
Calculate (-57 + (-30358)/(-516))/(-5*5/30).
-11/5
What is the value of 4582/632 + -5 + (-338)/8?
-40
Calculate 28475/(-2550) + 11 + 12 + -8.
23/6
Calculate 9 + -15 + (-69 - 68/4).
-92
What is ((-49855)/80)/13 - ((570/32)/(-3) - -6)?
-48
What is ((-3 - (-25)/3) + -10)*-6?
28
What is (-1)/((17 + 999/(-54))/(-33))?
-22
Calculate 28/112*-16 - ((-25)/1 - 0).
21
Evaluate (-68)/(-4) + (0 - 27730/1630).
-2/163
What is (4774/28116 - 18/108)*((-18)/8 - 0)?
-1/142
-31 + ((-1590)/(-120) - 18/(-24))
-17
Calculate (-1262)/180 - (-38)/(-342)*-63.
-1/90
What is the value of (-4)/(60/(-525))*(-6630)/1785?
-130
What is the value of (147 - (-2140179)/(-14535)) + (-2)/(-30)?
-3/17
What is the value of ((-24)/(-1360))/(21/14)?
1/85
What is the value of ((-102380)/1269512)/((-5)/(-6))?
-3/31
What is 1148/(-82)*3/18*-3?
7
What is the value of 406/(-22) + 28594/374 + -77?
-19
Calculate 18/8 - 1/(96/(-80))*318/(-60).
-13/6
Calculate 2/(-56) - 159736/109088.
-3/2
Calculate (-4)/7*(((-2560)/352)/(-80) - (-32)/11).
-12/7
(25/125*30/(-4))/(-13 - -1)
1/8
((9/(-155))/(-3))/(120756/(-402520))
-2/31
Calculate (1 - -1 - (-507)/(-208))*(482 - 484).
7/8
Calculate 20 - (-7552)/(-236)*((-1712)/232 - -8).
4/29
Evaluate -292*13/416 + -4*(-5)/160.
-9
What is (8961 + -8682)/(2 + 7)?
31
Evaluate -1 - ((-38)/95)/((84/(-45))/(210/(-35))).
2/7
Calculate (-89 + 40832/448)*(-56)/3.
-40
What is the value of (-47)/(36754/184) + (-2532)/(-68)?
37
(-112800)/4230*((-4)/(-6) + 2*(-7)/(-6))
-80
(12 - 1969/165)/((-14)/(-77)*(-6)/(-30))
11/6
Evaluate (8 + 116/(-13))/((-10626)/26312).
16/7
What is the value of 45 + 66/(-5346)*3627?
2/9
1/(1035/(-20010)*29/(-6))
4
What is the value of (99/(-1254))/((-34)/(-136))?
-6/19
What is 4/18*(297/5 - (-152 - -212))?
-2/15
What is ((-144)/(-30))/(2/(-3)*5220/(-63800))?
88
What is ((-7)/2 + 4)/(3208/16842)?
21/8
Evaluate ((-186)/(-3534) + (-898)/(-266))*(-21)/2.
-36
What is the value of (-7)/707 - (-773766)/20907?
37
What is (-42)/693 - ((-118773)/4455 + 22)?
23/5
Evaluate (3/(-3))/((-151)/755) - 18 - -84.
71
((-7714)/116)/(-19)*(-4)/(-903)*-3
-2/43
40 - (53 + -85 - -49)
23
Evaluate ((-5)/(-160) - (-26248)/(-4352)) + 78/15.
-4/5
Calculate 2553/(-888)*(-2 + 10).
-23
What is ((-57)/(-1))/(2270 - 2267)?
19
What is (-1)/(-95) - 73392184/(-1287820)?
57
What is (-31)/(-59334)*29 - 12010/(-132)?
91
Calculate (-2 - -16)/((-4)/33 + (12 - (-828)/(-66))).
-21
Calculate 456/(-380)*1*(-15)/9*(-58 - 1).
-118
What is the value of (490/343)/((-2448)/(-189) - 13)?
-30
Calculate (-7 + (-32776)/(-4692))*99/6 - 2/(-4).
6/23
Calculate ((-211)/1055)/((-144)/390) + -3 + (-7)/(-3).
-1/8
What is 1135/1816 - ((-1275)/(-312) - 3)?
-6/13
What is the value of 24*((-8)/6 + 0/(50 - 48))?
-32
What is the value of (26452/204 + -85)/((-6)/(-9))?
67
Calculate (-10621)/152 - (-180 + 110).
1/8
Evaluate (4/1)/((-180)/(-210))*(5 + -2).
14
What is the value of (-6 - -1)/((-10)/(-16)*(11 + 296/(-24)))?
6
((-24)/284)/(-114 - (-3234)/28)
-4/71
Evaluate 27*-2 - (-626460)/11660.
-3/11
What is the value of (124*299/333684)/(1/837)?
93
What is ((-1168)/(-141036))/(62/(-217))?
-2/69
Evaluate (((-35)/(-60))/7 + (-1393)/84)/(624/(-64)).
22/13
What is the value of 3528/43146 + 172/(-4386)?
2/47
Calculate 2 - (1421/1029 - (-33)/77).
4/21
What is the value of 10*-2 + -31 + (-3069)/(-66)?
-9/2
Evaluate (-159)/(-106)*((-36)/27)/(3 - 73/(-3)).
-3/41
What is the value of (-3 - -5)*-6 + (-5)/(65/(-247))?
7
What is ((-1280)/48)/(256/384)?
-40
Calculate (36 + -11 + 67)/(-4).
-23
What is (198/20)/(30/50)*(-754 - -752)?
-33
What is (92/(-690))/(1472/(-80) - -20)?
-1/12
What is 470/282*(0 - 0 - ((-21732)/(-1560) + -14))?
3/26
(-144)/180 - (12267/235 + 15/(-3)*1)
-48
4/(-8)*(15136/429 + (-98)/(-1911))
-53/3
What is the value of (855/(-375)*-10)/(288/360)?
57/2
What is the value of (-1312)/(-4674)*315/(-168)?
-10/19
Calculate 84 + -117 + 532/16.
1/4
What is (-6006)/147 + 12*12/(-126)?
-42
Calculate 1514/180166 - 59150/833.
-71
Calculate (-2859)/1722071*(-260)/(-30).
-2/139
What is the value of (1852/(-596) + 3)*(777/28 - 9) + 2?
-2/149
What is 4*(-2)/4 + (-32550)/5250?
-41/5
(5 + 3/3)*(-1022)/(-584)*(-24)/(-693)
4/11
What is (474/711)/(-3*33/(-27))?
2/11
What is (14/231)/(0 + (-6)/5) + (-262)/1179?
-3/11
What is the value of (19 + -109 - -128)/(-2)?
-19
What is the value of (-1)/(10032/1716 + 6/39) + (-129)/(-18)?
7
(-71)/(2556/(-2484)) + -31
38
What is the value of 6/((-150)/(-40)*(-2)/(-1))*-2?
-8/5
What is the value of (31785/585 - 82)/(((-4)/(-6))/(-2*1))?
83
What is (-57222)/1190 + -13 - -61?
-3/35
Calculate (-55 + 113 - 57)*10/(-8).
-5/4
What is the value of (-6)/15*980/882?
-4/9
What is the value of -2*27/420 + (-161)/(52969/(-141))?
3/10
What is 48/252 - 96798/484939?
-2/219
((-636)/1640 - -1 - 180/300)/((-3)/(-24))
4/41
5/15 - (-1918)/(-48) - (-183)/(-488)
-40
Evaluate (-870 - 10)/55 + 19.
3
What is (-17739)/405 + 5 + 82/(-10)?
-47
What is the value of ((-2431)/119 - -23)/(81/1890)?
60
((-3)/8*2)/(25673/(-1334996))
39
What is the value of (-9)/204*504/1134?
-1/51
Evaluate -51*3/63*(-29 - -36).
-17
Calculate ((-61)/(-96))/((-2414)/(-231744)).
61
66 + (-24)/96*188
19
What is the value of (32/(-33))/(17/(-357)) - 36/99?
20
Evaluate 8/(-7)*14/(-504)*42*18.
24
What is ((-1)/(8/(-12)))/(-4 + (-1652)/(-416))?
-52
Evaluate (50 - 52) + 11/(627/95).
-1/3
What is 195852/(-12885)*15/(-4)?
57
What is 2*2055/(-1800) + (28/(-48))/(-7)?
-11/5
Calculate ((-11)/(-2))/(729/4374).
33
Evaluate 66/(23562/3451) - (-2)/(-4)*54/3.
2/3
Evaluate 8/62 - 13580739/(-223107).
61
What is (135*31/(-5115))/(9/(-759))?
69
Calculate 5/25*((-10)/(-3) + -5)*69.
-23
Calculate -6*(904/(-48) + 6 - -11) - -13.
24
What is the value of (21*18/(-1890) + 244/(-30))*(-1 - 5)?
50
Calculate -18*18/(-1080) - (-4)/(400/(-355)).
-13/4
Evaluate ((-363)/(-2541))/((-10)/(-5)*(-2)/(-812)).
29
(-3591)/741*-29*37/(-7511)
-9/13
What is (((-31)/10)/(47/94))/(6/30)?
-31
What is the value of (-22)/(-11) + -41 - -35?
-4
What is 2/(-47) - 1094238/(-3938271)?
4/17
((21/(-196))/((-1440)/140))/(5/(-45))
-3/32
(-25)/3*(-5562)/(-1545)
-30
What is (402/(-1474))/((-1)/11)*5/51?
5/17
What is 7 - 213/994*(158/12 - -2)?
15/4
Evaluate (3/4)/(17985/30520).
14/11
((92 + -38)/(-9) - 180/(-14))*1/3
16/7
What is (-1248)/(-39) - -99*(-319)/1044?
7/4
Evaluate 2 - (-4)/(-6)*(-1536)/32.
34
What is 477/(-66) + (-41986)/(-5998)?
-5/22
Evaluate 9 - 1*592/(-16) - 12.
34
Calculate ((-7488)/819)/((-41)/(-287)).
-64
What is 2*(24/(-66))/((-256)/5984)?
17
What is the value of ((-22)/(-8))/(5*484/160)?
2/11
Calculate (23 + (-44205)/1925)/((-2)/(-20)).
4/11
What is the value of 26*26/2028*-363 - -40?
-81
1585/7608*24/2800*-5
-1/112
((-36704)/(-1036) - 36)*14/220
-2/55
Calculate 531/(-75) + ((-3)/(-2))/(28/(12936/99)).
-2/25
(6/5)/((-150)/12*6/(-5))
2/25
Evaluate ((-18)/14)/(-28 - 5850/(-210)).
9
What is (1545/(-206))/((1 - -8)/(-18) - 0)?
15
What is the value of (-3 - -2) + 3 - (-21 + -5 + -22 - -22)?
28
((-270)/(-75))/(26/30 - (-113 + 114))
-27
What is the value of 2982/(-213)*36/(-1428)?
6/17
What is (-315)/1995 + (-23)/((-4370)/16940)?
89
Calculate ((-40)/(-35))/(-4) - (-96782)/(-22785) - -5.
7/15
What is ((-28)/(-2))/(34/(-10200)*-175)?
24
-15 + 35365/2365 - (1856/(-774))/(-1)
-22/9
What is 146248/(-6969) - 45/3105?
-21
Calculate (-9 - -3 - -61)*(-22)/(-6 - 16).
55
Evaluate (-102)/(-16) + (26 + -39 - (-3 - 332/32)).
27/4
(30/(-54))/((-7260)/(-1089))
-1/12
What is the value of (-10 + (-31)/(-3))/((-36)/4428)?
-41
(-80)/(5*(-1938)/9180 - (-4)/18)
96
What is the value of 56/(-192)*4*(2 - 85/20)?
21/8
Evaluate (-216)/3348 - (-352)/(-310).
-6/5
Calculate 45 + (-95)/(5035/1166).
23
(-1150)/100 - ((-15)/(-6) - 4 - -1)
-11
What is the value of (-3)/1*((-1016064)/20412)/((-14)/6)?
-64
What is the value of (((-16)/24)/((-4)/(-18)))/(43/(-258)*-84)?
-3/14
What is the value of 2/((186/(-1085))/((66/56)/(-11)))?
5/4
What is (-5)/((-5)/(-6)) - 55440/(-13068)*3/2?
4/11
Calculate -20 - 3072/(-77 - -45) - -28.
104
What is 4/110*(829 - 109) - (-11 - (-246)/22)?
26
Evaluate (-8)/(-18) - (6134/(-1548) + (-342)/(-76)).
-4/43
What is the value of (50/((-1150)/(-69))*(-902)/33)/(1 + 0)?
-82
(16/64 + (-6)/8)*(-1600)/2400
1/3
What is the value of (-6)/(-13 - (-201)/15)*2/4?
-15
|
----- Forwarded by David M Gagliardi/TTG/HouInd on 11/08/01 08:45 AM -----
"Gagliardi,
Michael" To: "'[email protected]'" <[email protected]>,
<mgagliardi@un "'[email protected]'" <[email protected]>,
ocal.com> "'[email protected]'" <[email protected]>
cc:
11/08/01 06:59 Subject: FW: True Orange E-Mail Fax #109
AM
> -----Original Message-----
> From: [email protected] [SMTP:[email protected]]
> Sent: Wednesday, November 07, 2001 10:52 PM
> To: [email protected]
> Subject: True Orange E-Mail Fax #109
>
> True Orange E-Mail/Fax Service
> Volume 9, E-Mail/Fax #109, Wednesday, November 7, 2001
> Jerry Scarbrough's True Orange, P. O. Box 26530, Austin, Texas 78755 -
> Phone
> 512-795-8536
>
> At Least Five Top Recruits Coming to Austin This Weekend
>
> Five top recruits have confirmed they are coming to Austin this weekend
> and
> at least three of them are coming on official visits.
> The three who will be taking official visits to Texas are all
linebackers.
>
> They are Rufus Alexander, 6-2, 215, 4.49, of Baton Rouge Christian Life;
> A.
> J. Nicholson, 6-2, 240, 4.6, of Winston Salem Mount Tabor, and Michael
> Williams, 6-3, 220, 4.6, of Lindale.
> Alexander is one of the top recruits in Louisiana, Nicholson is one of
> North
> Carolina's top blue-chippers and Williams is emerging as one of the most
> coveted recruits in Texas.
> Two other national top 100 recruits will be in Austin this weekend, and
> one
> of them, TE Eric Winston, 6-6, 275, 4.9, of Midland Lee, also might be on
> an
> official visit. He is considering taking his official visit the Dec. 14
> weekend so he can attend the UT football banquet. If he decides to visit
> officially that weekend, his trip to Austin this weekend would be an
> unofficial visit.
> The other blue-chipper who confirmed he is coming to the UT-Kansas game
is
> QB
> Vincent Young, 6-5, 200, 4.5, of Houston Madison.
> Young is as good a quarterback prospect as there is in the nation. Texas
> and
> Miami are his top two schools right now. Winston also is considering
Texas
>
> and Miami, along with Texas A&M.
> Alexander, who has scholarship offers from most of the Southeastern
> Conference powers, is leaning toward Texas, but also is considering
> Alabama,
> Tennessee and Florida.
> Nicholson committed to North Carolina during the summer, but decommitted
> when
> the Tar Heels lost their first three games. He is considering Texas,
North
>
> Carolina and several other ACC and SEC schools.
> Williams committed to Oklahoma on a visit about three weeks ago, but then
> decommitted and visited Texas A&M and Texas Tech. He had planned to visit
> Texas the Dec. 14 weekend, but told me tonight he decided to take his
> official visit now because "I'm getting kind of tired of recruiting." He
> said
> he expects to make a decision soon, and said no one has an edge right
now.
> The Longhorns also are trying to get Lance Mitchell, a 6-2, 240-pound
> inside
> linebacker from City College of San Francisco to visit. He runs under 4.6
> in
> the 40 and is one of the top junior college prospects in the country. He
> said
> he is considering visiting Texas, but has not set a date.
> Texas has 16 commitments right now, and LB Garnet Smith of Arlington
Lamar
>
> now says he is a solid UT pledge. "I'm going to sign with Texas," he
said.
> He
> has committed at various times to Texas, Oklahoma and Tennessee.
> * * * *
> Coach Mack Brown said Wednesday the Longhorns have been practicing very
> hard
> this week. He said the team has great senior leadership and has come to
> realize that top five teams have to show up ready to practice hard every
> day
> and play hard every Saturday.
> He also said OG Antwan Kirk-Hughes, who missed the last two games with a
> sprained ankle, should be able to play against Kansas. "He worked out the
> last two days, and we think he'll be able to play." Brown said Beau Baker
> played at a very high level when he was called upon at Baylor when
Tillman
>
> Holloway suffered a shoulder injury.
> Holloway won't play against Kansas because of a sore shoulder and will be
> examined again next week to determine if he can resume practicing.
> With Derrick Dockery and Baker the only completely healthy front-line
> guards,
> Brown said center Matt Anderson is ready to play at guard if needed, and
> Jason Glynn could move in at center.
> "It seems like we have one position where we get a lot of injuries each
> year," he said. "Last year, it was the tight ends who got hurt. This
year,
>
> it's the offensive guards."
> * * * *
> My next e-mail/fax will be whenever events warrant.
> * * * *
> The True Orange E-Mail/Fax Service includes at least 99 fax/e-mails a
> year and costs $99 ($79 by E-Mail). The True Orange Newsletter includes
> 26
> newsletters and is published weekly during football season and twice
> monthly
> during most of the other months. It costs $45. Save by subscribing to
both
>
> for $130 (or $110 if you take the faxes via E-Mail or $99 if you take
> both
> services via E-Mail). Send check to address at the top of page. I also
> update my 900 number - 1-900-288-8839 - daily with recruiting news. My
> E-Mail address is: [email protected].
|
Nespresso Machine
Buying Guide for the Best Nespresso Machine
Joan Mackenzie
08/29/2018 07:10PM
With the many electronic galore, in the modern world, it is very difficult to find a good Nespresso machine that will serve you for a long time. You need to carry out a careful research to help you determine the right machine for you. You all know that Nespresso machine is normally very convenient but depending on the kind of manufacturer that you are choosing, it is important for you to look out for a number of things. Here are some of the things that you will need to ensure that you watch out for especially if you would to consider quality in your business.
You need to ensure that you choose a coffee flavor that will work for you in the right manner. You would like to have that Nespresso machine that will make the coffee that you have always wanted in life. In the case, for instance, you are a fan of cappuccinos, you may always want to ensure that you get a Nespresso machine that will make the milk steamer as required in the right manner. There is always the need to ensure that you get the best nespresso machine, that is compatible with coffee capsule I the right manner.
The design and the size of the Nespresso machine is one thing that you need to ensure that you consider in the right manner. We live in an era whereby machines will come in different designs and sizes that may suit the needs that you have in the right manner. You will need to ensure that you choose a size that will ensure that you are able to get the right strategies in the best way. A machine that is compatible with the place that you are placing it if it is your office or your living room so that you have an easy way of operating it.
The features of your Nespresso machine is another thing that you need to put in place. You find that there are machines that will make more than one coffee at the same time, choose one that will make you have the right services in the right manner. The cup size of the Nespresso machine that you choose is also very important, you need to ensure that you are able to get the best to ensure that you have the right facilities as you have always wanted in the right manner. Take your time to ensure that you get a strategy that is suitable for you in the right manner. Find out more here: https://www.huffingtonpost.com/2014/09/29/how-to-clean-coffee-maker_n_5861026.html.
|
Chiplun railway station
Chiplun railway station is a train station on Konkan Railways. It is located just off NH 66 in the Konkan town of Chiplun. The preceding station on the line is Anjani railway station and the next station is Kamathe railway station.
Trains
10103/10104 Mandovi Express
10111/10112 Konkan Kanya Express
11003/11004 Tutari Express
11085/11086 Lokmanya Tilak Terminus - Madgaon AC Double Decker Express
11099/11100 Lokmanya Tilak Terminus - Madgaon AC Double Decker Express
12051/12052 Dadar Madgaon JanShatabdi Express
12483/12484 Kochuveli - Amritsar Weekly Express
12617/12618 Mangala Lakshadweep Express
12619/12620 Matsyagandha Express
12741/12742 Vasco da Gama - Patna Superfast Express
16335/16336 Gandhidham−Nagercoil Express
16345/16346 Netravati Express
19259/19260 Kochuveli−Bhavnagar Express
19331/19332 Kochuveli - Indore Weekly Express
22113/22114 Lokmanya Tilak Terminus−Kochuveli Express
22115/22116 Lokmanya Tilak Terminus - Karmali AC Superfast Express
22119/22120 Mumbai CSMT - Karmali Tejas Express
22149/22150 Pune–Ernakulam Express
22475/22476 Hisar - Coimbatore SF AC Express
22629/22630 Dadar-Tirunelveli Express
22633/22634 Thiruvananthapuram - Hazrat Nizamuddin Express
50103/50104 Dadar Central - Ratnagiri Passenger
50105/50106 Sindhudurg Passenger
References
Category:Railway stations along Konkan Railway line
Category:Railway stations in Ratnagiri district
Category:Ratnagiri railway division
|
One month ago, Democrats viewed Arizona and Georgia as winnable states for the first time in decades. And demographic trends seem to have put Texas in play as well.
But then came Hillary Clinton’s shocking loss to Donald Trump, powered largely by her unexpected weakness across the traditionally Democratic industrial states of the Rust Belt.
Now, party officials in the fast-growing, so-called purple states that so recently looked within reach — all of which also fell to Trump — are increasingly wary that national party leaders will redirect their focus toward Michigan, Pennsylvania and Wisconsin, denying them the critical attention and resources that might otherwise accelerate their movement toward swing-state status.
“Right now, Democrats need to make sure we correct but we don’t overcorrect,” said Atlanta Mayor Kasim Reed. “This isn’t a situation where we need to be making false choices. We need to understand we won the popular vote, [but] we lost states we should not have lost."
The worry is at the heart of the quandary the party now faces as it attempts to find a middle ground between the coalition that elected Barack Obama to two terms but failed to materialize for Clinton, and the suddenly ascendant one that elected Trump to the presidency.
Current and aspiring leaders of the party have in recent days repeatedly pointed to the importance of keeping white working-class voters in the Democratic column, speaking of the need to reconfigure the economic message coming from their standard-bearers up and down the ballot so that it appeals to all voters, not just the African-American and Hispanic constituencies that make up the party base.
That’s been a top stated concern of both Vermont Sen. Bernie Sanders and Massachusetts Sen. Elizabeth Warren as they seek to map out the party’s future. Ohio Sen. Sherrod Brown kicked off the week with a New York Times op-ed making a similar case.
Yet Democrats in places like Arizona, Georgia and Texas — states where growing minority populations hold long-term promise for the party — fear the reaction to Clinton’s loss could lead to a zero-sum approach where the rush to win back wayward white voters in the Rust Belt comes at their expense.
"I do have this worry where there's going to be an opportunity-cost scenario, where they say we know how to win in these states, but not these others,” says Arizona Democratic Rep. Ruben Gallego. “But the thing I like to point out to people is the same working-class people in Michigan are in Arizona. They vote the same, they may just not live in the same geography."
Democrats in Arizona and Georgia are quick to note that the party is as close as it’s ever been to being competitive at the presidential level in those states. On the strength of a surge in Latino voters, Clinton lost Arizona — where the GOP nominee has won in 16 of the past 17 elections — by less than 4 percentage points. In Georgia, Clinton lost by just 5 points as African-American voters turned out for Clinton.
Those results came despite minimal national investment in those two states. The Clinton campaign and its supportive super PAC ended up pumping ad money into both toward the end of the race — and Clinton herself visited Arizona once — but the vast majority of her team’s time and resources went toward core swing states.
“What was going on in Georgia is real. It was the most competitive presidential election in Georgia since 1992, and we did it with minimal resources,” said Reed. “We’re going to continue to do our work here and grind it out, but for the national Democratic Party it’s really important for us — we need to win Southern states going forward. And it sends a message even to the Rust Belt and Midwestern communities [that] when you bring along Southern states, it helps us maintain our status as a national party."
The Arizona and Georgia margins turned out to be closer than in Ohio and Iowa — two swing states Obama won twice — and Trump became the first Republican since Ronald Reagan to win Pennsylvania, Michigan and Wisconsin in the same election. As such, much of the public conversation surrounding the party’s post-Clinton future has zeroed in on what went wrong in those traditional battlegrounds.
“Our national committees are not the most forward-looking institutions. The way these committees work is: it’s better to fail doing a conventional thing rather than trying something new,” said Andy Barr, a Democratic strategist in Arizona. “We’ve lost a ton of voters in the Midwest. Maybe they’ll come back, I don’t know, but those voters haven’t voted for any Democrat other than Barack Obama in a long freaking time. But with minimal investment in Arizona and Texas, those are the states where we gained voters, with almost no involvement from D.C."
Nonetheless, the party is unlikely to undergo any significant reallocation of resources for at least a few months, while it goes through a protracted period of soul-searching and elects a new Democratic National Committee chairman who will help direct such decisions.
“The conversation that people have in the initial aftermath of a shocking defeat is hopefully different than the one that we will have when we’re really mapping out the future,” said Seth Scott, an aide to Phoenix’s mayor who served as Clinton’s Arizona state director, and who said the state would be crucial to any Democrat’s White House hopes in 2020. “The last two weeks have been really difficult, not only for all of us personally, but for a party that is trying to grapple with not only how we lost, but how we lost to the person we lost to, and the way we lost."
And in the meantime, the party will go through a series of autopsies of what went wrong in November, a process that will likely include more rounds of second-guessing about the Clinton campaign and national Democrats’ money distribution.
"Look at what happened here in Texas without heavy investment or infrastructure. We won four of the five big metro areas in Texas," said Texas Democratic Rep. Henry Cuellar. “Whoever the new DNC chair's going to be, we're hoping that there will be an emphasis on areas that will have growth opportunities, and Texas is one of them. But we're already not getting those resources. So are they taking anything away from us? No, we're not getting it already."
Reed, the Atlanta mayor, makes the point that Clinton lost North Carolina and its 15 electoral votes — where her team spent tens of millions of dollars — by just roughly 1 point less than she lost Georgia and its 16 votes.
“It would be a mistake to not look at the gains that were made in Georgia,” said Rebecca DeHart, executive director of that state’s Democratic Party, nodding to the looming uncertainty about resources.
“It’s the million-dollar question. Literally."
|
Look at the various Mailing
Lists about Zope. The Mailing Lists are where you can get
quick, accurate, friendly help from a large community of Zope
users from around the world.
Browse and search the integrated,
Online Help System which contains documentation on
the various kinds of components you'll find in Zope.
Go directly to the
Zope Management Interface if you'd like to start working with Zope
right away. NOTE: Some versions of Microsoft Internet Explorer,
(specifically IE 5.01 and early versions of IE 5.5) may have problems
displaying Zope management pages. If you cannot view the management pages,
try upgrading your IE installation to the latest release version, or use
a different browser.
|
DC Bombshells Wonder Woman Chase Dorbz Vinyl Figure #414
Product Description
$14.99
Maximum quantity available reached.
It's up to the woman of DC to save the day and they're doing it in style! Combining the popular pin-up art style of Ant Lucia with a Funko twist, the DC Bombshells join the line-up of Dorbz Vinyl Figures. Packaged in a window display box, this DC Bombshells Wonder Woman Dorbz Vinyl Figure #414 measures approximately 3-inches tall.
|
Follow NewLeveller on Twitter
Category: Syria
Suppose you are an arms manufacturer (you may actually be an arms manufacturer as this is a public post!). There are three ways of making money from your products: planned or unplanned obsolescence; high maintenance costs; get your clients to buy more product, perhaps by actually using them.
Although wars have traditionally been the quickest driver of new weapons (think radar, tank, jet engine and atomic bomb as twentieth century examples) the current pace of technology makes obsolescence a realistic short-term possibility. Suppose, however, you are Lockheed Martin who make the Trident system. It is clear that a ballistic missile maker won’t get a great many sales! Its like buying a car which is too reliable and lasts too long. So, as with the car, tying the customer to a maintenance schedule is important. Along with the Royal Navy submarine upkeep, the annual maintenance cost of Trident is estimated to be about £2.3bn.
What about the most destructive way of making money, getting your customers to use their weapons? Considering the wars that have taken place over the past twenty five years it is possible to work out the desirable parameters. The war must be large enough to necessitate the use of expensive, probably missile, technology (and the occasional destruction of an asset such as a fighter plane) Trump firing 59 Cruise missiles at Syria at around $1mil per shot is perfect! And that was just a simple retaliatory action. Small ‘boots on the ground’/police actions with mostly rifles and bullets aren’t going t make that much profit (as in, for example, saving innocents from a central African genocide). Besides, open spaces like deserts really allow you to use the expensive stuff!
But the war must not be too big that it destabilises the world. Fellow large multi-national Corporations like stability, so the war should not really disrupt world trade to any extent; profit forecasts must be met. Besides, ending the world is very bad for repeat sales (even for Lockheed Martin!). So the West is very cautious in North Korea!
Next the war must not be seen to have too many civilian casualties or the casualties must be easily presented as evil and thus disposable. Ideally there should be few or none of your own casualties. This means, either way, the public will not get on the politicians backs to stop the conflict.
Finally the war, either during its execution but most certainly afterwards must present your investors with opportunities for profits from reconstruction. A few months ago I wrote a post about the situation in Syria and included a report from the Washington Post about major Gulf States (including Saudi Arabia) financial backing for Assad. If we had really wanted to sort out Syria then this is where we should have started, with the Middle East banks. But there is no profit in this, unlike arms to be used in Yemen!
So to sum up, Small proxy wars are best just as long as the controlling powers understand the situation and it does not get out of control.
Maybe this post is simply cynical. But allowing for the fact that oil is an important driving factor in when and where we decide to go to war we are always ready to defend liberty where this a private profit to be made for organisations such as BAE systems. Otherwise, you are on your own. However cynical this post, it is nowhere near the level shown by arms manufacturers!
I have never held a gun, let alone fired one. I share this with the vast majority of British people, lucky to be born in the second half of the 20th Century when being sent to war in a mass army was a thing of the past. At the same time I am no idealistic dreamer and am fully aware that we live in a world full of dangers (albeit some of which we create!), appreciating that there are men and women who make sacrifices for our country. Because I have never been in the forces I cannot fully understand the life of a serviceman/woman, but I am quite capable of questioning the motives of a British establishment which commits them to action. In particular there is a serious issue with a royal family which treats the armed forces both as a mean of personal glorification and a job creation scheme.
One of the inspirational aspects of the past few months for me has been meeting members of the Veterans for Peace movement (I particularly valued their presence at Levellers Day and Gus Hales has written a personal account). A number of my recent blogs (here for example) have involved the way contemporary monarchy encourages many people in Britain to uncritically accept it as part of their identity. For servicemen and women the pressure must be overwhelming with the taking of the oath of loyalty and the justification of fighting for King/Queen and Country. To mentally reject that identity and question whether military power is in the interests of the British people themselves takes real will power.
While limiting the problem of the glorification of war to the monarchy risks missing a large part of the story, it is still a good place to start. The fact that for three or four hundred years following the Norman Conquest English monarchs were in reality successful warlords means that monarchy and militarism were interlinked from the start. Although by the eighteenth century the time was long past when a king personally led an army, monarchs lost none of their zeal for sending troops into battle for power and glory. For example in 1781 when it was patently clear to politicians that the war in North America was lost and British troops should be withdrawn, George III (pictured above) insisted on continuing with hostilities, With an increasingly rebellious House of Commons, Prime Minister Lord North was left with no option but to tell the King where to go! Today, George III’s ancestors show similar disdain for servicemen. At any state occasion members of the Royal family can be seen ridiculously strutting around in uniforms of high military rank bedecked with ribbons and medals.
As this Telegraph article makes clear, many of these are invented or handed out by the queen presumably with the intention of making an impression on us. There can be no other reason why you would give Prince Philip the Order of Merit when it is limited to 24 individuals and otherwise has been held by such luminaries as Bertrand Russell (himself a pacifist!). Look more closely and you will see almost all of these preening people displaying medals such as the Queens Silver Jubilee medal and Golden Jubilee medal given ‘for service’. Interestingly these very medals have frequently been denied or withheld from actual servicemen/women of long standing as this blog illustrates.
|
Q:
get cloudformation_facts for multiple cloudformation stacks with with_items in ansible
I have multiple clouformation stacks, and stored their names as a list in CF_TEMPLATE_ITEMS
Now I am trying to gather information about all of them (at the end I want to have the stack_output of all of them):
- name: Get all facts for all cf stacks
cloudformation_facts:
stack_name: "{{ item }}"
with_items: "{{ CF_TEMPLATE_ITEMS }}"
Unfortunately, afterwards cloudformation contains only the information for the last stack. It seems the information for the others are overwritten.
Can I somehow cat the facts for all cloudformation stacks from of list of stack names?
A:
Yes, cloudformation_facts overwrites cloudformation fact with every run.
To collect data from every run, register loop result and reformat it to a clean dict, like this:
- cloudformation_facts:
stack_name: "{{ item }}"
with_items: "{{ CF_TEMPLATE_ITEMS }}"
register: cf_tmp
- set_fact:
cf: "{{ dict(cf_tmp.results | map(attribute='ansible_facts.cloudformation') | map('dictsort') | sum(start=[])) }}"
This code is not tested. This should give you cf dict with all your stacks facts as its keys.
|
This article is part of our The Armband series.
Maybe I didn't study enough for the last MLS slate. Maybe it was a regular MLS week. Either way, most of the point leaders were players who weren't captained, and the only one in my top five who managed double-digit fantasy points was Maximiliano Moralez. Otherwise, David Villa left early with injury, while Miguel Almiron and Sebastian Giovinco, usually good for bonus points, failed to do anything outside of that. I pointed out Borek Dockal in my article but never considered him for the armband, and he led the week with 17 fantasy points. That's kind of how things work. Sometimes you step outside of the box and it works, but most of the time it doesn't.
For the upcoming week, there is one team that reigns over the rest: Atlanta, who are at home against Orlando City, a side that has lost their last seven matches while allowing multiple goals in each. Atlanta only scored once last match, but a lot of that can be attributed to a difficult opponent. This opponent could lead to a huge week for Almiron and hat-trick maestro Josef Martinez. The downside to Almiron is that he's surpassed seven fantasy points once in his last eight matches. The downside to Martinez is the three points he had last week, but the upside is his MLS hat-trick record.
Unfortunately, there aren't many other matchups that stand out. Nicolas Lodeiro is always good for a few bonus points, especially at home. Maybe the best thing for Seattle is that Portland was in Atlanta last weekend and may have tired legs with another difficult matchup. Then again, it's a rivalry and it'll be a physical match. Ignacio Piatti is at home, but trusting him against the Kansas City back line doesn't feel like a good idea. As for Moralez and Dockal, both are on the road against difficult competition. I'd recommend the Earthquakes, but they're up against the Galaxy, who have allowed only nine goals in seven road matches.
The next best place to look is Columbus at home against Real Salt Lake. I'm not comfortable with a clean sheet, but Federico Higuain and Gyasi Zardes should be good for a goal or two. Outside of the Atlanta match, both have been great at home, securing at least eight fantasy points in the three contests prior to hosting Atlanta. RSL have just four points from eight road matches this season, scoring five times and giving up 20.
New England are home to D.C. United, and that puts Diego Fagundez back on the radar after scoring at least eight fantasy points in the last three matches. He's more of a differential play, but he's quietly remained fantasy relevant for most of the season with five goals and five assists. Teal Bunbury has nine goals, if you're into that sort of thing.
If you want to wait for Sunday's matches and hope for the best, there are a few options. Toronto are at home, but they've been all over the place, scoring (and allowing) four against D.C. a couple weeks ago and then going scoreless against Dallas before that. The Red Bulls are one of the few teams with a positive road goal differential, which doesn't help. Giovinco hasn't provided much of a fantasy spark this season, similar to Toronto as a whole, yet he's always in the running for the armband. The safer play may be Victor Vazquez, who has at least eight fantasy points in his last four matches, scoring three and assisting four.
Elsewhere on Sunday, the Whitecaps are only in play because they host the Rapids, who have scored two and allowed 13 on the road. Felipe Martins has the best floor, but Kei Kamara scored a brace in the last home match and Cristian Techera had a hat trick in the one before that. If you like Vancouver, go ahead and guess the goal scorer, but I'll pass for my captain.
The ultimate differential would be Adama Diomande, who has filled in for Carlos Vela as LAFC's best player with four goals in the last three matches. The Union have a negative-nine goal differential on the road, so he's definitely viable.
|
Dr. Justine Kolker is currently in the Dentist-Scientist program at The University of Iowa College of Dentist. She is presently continuing her clinical specialty in Operative Dentistry and her basic science PhD in Oral Science (Public Health). Dr. Kolker is still in the introductory phase of her research examining the failure/survival rats after large amaigams or crowns are placed. A second phase will include reviewing records of randomly selected patients from the original target group. Auditing charts, examining radiographs, and using computerized data will be done to ascertain patient-and tooth level variables of interest. The final study may include analyses data obtained from a pubic (Medicaid, or private source (Blue Cross/Blue Shield, Private Practice Network). Utility opportunity costs, cost benefits, and patient perspectives are other factors that may be analyzed. Dr. Kolker attended the Academy of Operative Dentistry Annual meeting in Chicago, IL 2/18-2/21/98.
|
Q:
VBA: Skip cells in range
I'm trying to fill formulas in column C like this:
LastRow = Range("A65536").End(xlUp).Row
Range(Cells(1, 3), Cells(LastRow, 3)).Formula = "=A1*B1"
And it works fine. But is it possible to skip those cells in column C, which value is for example > 0 ?
A B C -> A B C
3 2 0 3 2 6
3 4 0 3 4 12
5 6 5 5 6 5 <- this value is not updated
A:
For this you will have to do a slight change to how your data looks like. For example, you need to add "Header" to the first row. The reason we would do that is because we would be using AutoFilter
Let's say your data looks like this
Now use this code
Sub Sample()
Dim ws As Worksheet
Dim copyFrom As Range
Dim lRow As Long
'~~> Change this to the relevant sheet
Set ws = ThisWorkbook.Worksheets("Sheet1")
With ws
'~~> Remove any filters
.AutoFilterMode = False
'~~> Get lastrow in column c
lRow = .Range("C" & .Rows.Count).End(xlUp).Row
'~~> Filter the col C to get values which have 0
With .Range("C1:C" & lRow)
.AutoFilter Field:=1, Criteria1:="=0"
Set copyFrom = .Offset(1, 0).SpecialCells(xlCellTypeVisible)
'~~> Assign the formula to all the cells in one go
If Not copyFrom Is Nothing Then _
copyFrom.Formula = "=A" & copyFrom.Row & "*B" & copyFrom.Row
End With
'~~> Remove any filters
.AutoFilterMode = False
End With
End Sub
Output
A:
One way to do it through loop :
Sub test()
Dim rngTest As Range
Dim rngCell As Range
LastRow = Range("A65536").End(xlUp).Row
Set rngTest = Range(Cells(1, 3), Cells(LastRow, 3))
Application.Calculation = xlCalculationManual
For Each rngCell In rngTest.Cells
If Not rngCell <> "" Then
rngCell.Formula = "=" & rngCell.Offset(, -2).Address & "*" & rngCell.Offset(, -1).Address
End If
Next
Application.Calculation = xlCalculationAutomatic
End Sub
|
A jury ruled George Zimmerman “not guilty” tonight, but some enraged Twitter users are threatening to take the law into their own hands:
https://twitter.com/Tyrin___/status/356232215149023234
https://twitter.com/WiseGuyWilliams/status/356235082421374976
https://twitter.com/pretty_boyqotti/status/356232347097640962
Put this on mama I will kill Zimmerman — December25 (@Trill_IsBack23) July 14, 2013
I Hope Somebody Kill Zimmerman Tonight!!! — Nickole.. (@EfWatU_TlknBout) July 14, 2013
SUMBODY MUST KILL ZIMMERMAN ASAP!!!! — $$$FAT TREL$$$ (@FATTREL) July 14, 2013
https://twitter.com/RealDeal_LiL/status/356233854228512770
https://twitter.com/DntTkMeSerious/status/356231856053686273
https://twitter.com/Schwag_OD/status/356233816890810369
https://twitter.com/MikeLewwDaPLUG/status/356233539932532739
Note the Twitter Rules:
|
Woodchuck Cider
Categories
Farm/Food Artisan
Description
First crafted in Proctorsville, Vermont in 1991, Woodchuck Hard Cider now operates out of Middlebury, Vermont. Partnering with businesses such as Champlain and Sunrise Orchards, Woodchuck Hard Cider is committed to quality and fostering local relationships with Vermont apple growers. As the number one selling hard cider in the country and distributed nationwide, Woodchuck Hard Cider works hard to utilize as much local juice as is available.
|
[Effects of endotoxin/lipopolysaccharide on proliferation and apoptosis of human umbilical cord mesenchymal stem cells].
To investigate the effects of different concentrations of lipopolysaccharide (LPS) on proliferation and apoptosis of human umbilical cord mesenchymal stem cells (hUCMSCs) in vitro, and to explore their possible mechanism. hUCMSCs from umbilical cord tissue of full-term healthy fetus delivered by caesarean section were isolated and cultured in vitro using tissue attachment method. The 3rd passage hUCMSCs were used in the study. Cells were divided into groups A, B, C, D, and E, which were treated with DMEM/F12 medium containing 0, 0.1, 1.0, 10.0, and 100.0 µg/mL of LPS respectively. In groups B, C, D, and E, methyl-thiazole-tetrazolium assay was used to detect proliferative activity of hUCMSCs at post treatment hour (PTH) 12, 24, and 48 (denoted as absorption value), with 5 samples in each group at each time point; apoptosis of hUCMSCs at PBH 24 was identified with acridine orange-ethidium bromide (AO-EB) staining, with 4 samples in each group; apoptotic rate of hUCMSCs was determined by flow cytometer, with 5 samples in each group. Above-mentioned indexes were determined in group A at the same time points. Data were processed with analysis of variance and LSD- t test. (1) There was no statistically significant difference in proliferative activity of hUCMSCs at PTH 12 among groups A, B, C, D, and E (with t values from -1.67 to 1.33, P values above 0.05). Compared with that of group A, proliferative activity of hUCMSCs was increased in groups B, C, and D at PTH 24 and 48 (with t values from -13.42 to 17.34, P < 0.05 or P < 0.01), especially so in group C. Proliferative activity of hUCMSCs was lower in group E at PTH 24 and 48 than in group A (with t values respectively 8.64 and 17.34, P values below 0.01). (2) Obvious apoptosis of hUCMSCs was observed in group E but not in the other 4 groups with AO-EB staining. (3) Apoptosis rates of hUCMSCs in groups A, B, C, D, and E were respectively (3.1 ± 0.6)%, (2.6 ± 0.7)%, (2.9 ± 0.8)%, (3.1 ± 0.4)%, (25.1 ± 2.7)% (F = 272.19, P < 0.01). Apoptotic rate of hUCMSCs in group B, C, or D was respectively close to that in group A (with t values respectively 1.22, 0.57, -0.14, P values above 0.05), but it was higher in group E than in group A (t = -17.63, P < 0.01). hUCMSCs proliferation may be promoted by low concentration of LPS. hUCMSCs proliferation is inhibited or induced to apoptosis along with the increase in concentration of LPS, and it may be related to activation of different major molecular signaling pathways by different concentrations of LPS.
|
<?xml version="1.0" encoding="UTF-8"?>
<?import java.lang.*?>
<?import javafx.scene.control.*?>
<?import javafx.scene.layout.*?>
<VBox prefHeight="400.0" prefWidth="600.0" xmlns="http://javafx.com/javafx/8" xmlns:fx="http://javafx.com/fxml/1">
<children>
<TableView prefHeight="200.0" prefWidth="200.0">
<columns>
<TableColumn fx:id="rt36633_tableColumn" prefWidth="599.0" text="tableColumn1" />
</columns>
</TableView>
<TabPane>
<tabs>
<Tab fx:id="rt36633_tab" text="tab1" />
</tabs>
</TabPane>
<MenuBar>
<menus>
<Menu fx:id="rt36633_menu" text="File">
<items>
<MenuItem fx:id="rt36633_menuItem" text="Close" />
</items>
</Menu>
</menus>
</MenuBar>
</children>
</VBox>
|
Q:
Restoring a Tensorflow model that uses Iterators
I have a model that's trains my network using an Iterator; following the new Dataset API pipeline model that's now recommended by Google.
I read tfrecord files, feed data to the network, train nicely, and all is going well, I save my model in the end of the training so I can run Inference on it later. A simplified version of the code is as following:
""" Training and saving """
training_dataset = tf.contrib.data.TFRecordDataset(training_record)
training_dataset = training_dataset.map(ds._path_records_parser)
training_dataset = training_dataset.batch(BATCH_SIZE)
with tf.name_scope("iterators"):
training_iterator = Iterator.from_structure(training_dataset.output_types, training_dataset.output_shapes)
next_training_element = training_iterator.get_next()
training_init_op = training_iterator.make_initializer(training_dataset)
def train(num_epochs):
# compute for the number of epochs
for e in range(1, num_epochs+1):
session.run(training_init_op) #initializing iterator here
while True:
try:
images, labels = session.run(next_training_element)
session.run(optimizer, feed_dict={x: images, y_true: labels})
except tf.errors.OutOfRangeError:
saver_name = './saved_models/ucf-model'
print("Finished Training Epoch {}".format(e))
break
""" Restoring """
# restoring the saved model and its variables
session = tf.Session()
saver = tf.train.import_meta_graph(r'saved_models\ucf-model.meta')
saver.restore(session, tf.train.latest_checkpoint('.\saved_models'))
graph = tf.get_default_graph()
# restoring relevant tensors/ops
accuracy = graph.get_tensor_by_name("accuracy/Mean:0") #the tensor that when evaluated returns the mean accuracy of the batch
testing_iterator = graph.get_operation_by_name("iterators/Iterator") #my iterator used in testing.
next_testing_element = graph.get_operation_by_name("iterators/IteratorGetNext") #the GetNext operator for my iterator
# loading my testing set tfrecords
testing_dataset = tf.contrib.data.TFRecordDataset(testing_record_path)
testing_dataset = testing_dataset.map(ds._path_records_parser, num_threads=4, output_buffer_size=BATCH_SIZE*20)
testing_dataset = testing_dataset.batch(BATCH_SIZE)
testing_init_op = testing_iterator.make_initializer(testing_dataset) #to initialize the dataset
with tf.Session() as session:
session.run(testing_init_op)
while True:
try:
images, labels = session.run(next_testing_element)
accuracy = session.run(accuracy, feed_dict={x: test_images, y_true: test_labels}) #error here, x, y_true not defined
except tf.errors.OutOfRangeError:
break
My problem is mainly when I restore the model. How to feed testing data to the network?
When I restore my Iterator using testing_iterator = graph.get_operation_by_name("iterators/Iterator"), next_testing_element = graph.get_operation_by_name("iterators/IteratorGetNext"), I get the following error:
GetNext() failed because the iterator has not been initialized. Ensure that you have run the initializer operation for this iterator before getting the next element.
So I did try to initialize my dataset using: testing_init_op = testing_iterator.make_initializer(testing_dataset)). I got this error: AttributeError: 'Operation' object has no attribute 'make_initializer'
Another issue is, since an iterator is being used, there's no need to use placeholders in the training_model, as an iterator feed data directly to the graph. But this way, how to restore my feed_dict keys in the 3rd to last line, when I feed data to the "accuracy" op?
EDIT: if someone could suggest a way to add placeholders between the Iterator and the network input, then I could try running the graph by evaluating the "accuracy" tensor while feeding data to the placeholders and ignoring the iterator altogether.
A:
When restoring a saved meta graph, you can restore the initialization operation with name and then use it again to initialize the input pipeline for inference.
That is, when creating the graph, you can do
dataset_init_op = iterator.make_initializer(dataset, name='dataset_init')
And then restore this operation by doing:
dataset_init_op = graph.get_operation_by_name('dataset_init')
Here is a self contained code snippet that compares results of a randomly initialized model before and after restoring.
Saving an Iterator
np.random.seed(42)
data = np.random.random([4, 4])
X = tf.placeholder(dtype=tf.float32, shape=[4, 4], name='X')
dataset = tf.data.Dataset.from_tensor_slices(X)
iterator = tf.data.Iterator.from_structure(dataset.output_types, dataset.output_shapes)
dataset_next_op = iterator.get_next()
# name the operation
dataset_init_op = iterator.make_initializer(dataset, name='dataset_init')
w = np.random.random([1, 4])
W = tf.Variable(w, name='W', dtype=tf.float32)
output = tf.multiply(W, dataset_next_op, name='output')
sess = tf.Session()
saver = tf.train.Saver()
sess.run(tf.global_variables_initializer())
sess.run(dataset_init_op, feed_dict={X:data})
while True:
try:
print(sess.run(output))
except tf.errors.OutOfRangeError:
saver.save(sess, 'tmp/', global_step=1002)
break
And then you can restore the same model for inference as follows:
Restoring saved iterator
np.random.seed(42)
data = np.random.random([4, 4])
tf.reset_default_graph()
sess = tf.Session()
saver = tf.train.import_meta_graph('tmp/-1002.meta')
ckpt = tf.train.get_checkpoint_state(os.path.dirname('tmp/checkpoint'))
saver.restore(sess, ckpt.model_checkpoint_path)
graph = tf.get_default_graph()
# Restore the init operation
dataset_init_op = graph.get_operation_by_name('dataset_init')
X = graph.get_tensor_by_name('X:0')
output = graph.get_tensor_by_name('output:0')
sess.run(dataset_init_op, feed_dict={X:data})
while True:
try:
print(sess.run(output))
except tf.errors.OutOfRangeError:
break
A:
I would suggest to use tf.contrib.data.make_saveable_from_iterator, which has been designed precisely for this purpose. It is much less verbose and does not require you to change existing code, in particular how you define your iterator.
Working example, when we save everything after step 5 has completed. Note how I don't even bother knowing what seed is used.
import tensorflow as tf
iterator = (
tf.data.Dataset.range(100)
.shuffle(10)
.make_one_shot_iterator())
batch = iterator.get_next(name='batch')
saveable_obj = tf.contrib.data.make_saveable_from_iterator(iterator)
tf.add_to_collection(tf.GraphKeys.SAVEABLE_OBJECTS, saveable_obj)
saver = tf.train.Saver()
with tf.Session() as sess:
tf.global_variables_initializer().run()
for step in range(10):
print('{}: {}'.format(step, sess.run(batch)))
if step == 5:
saver.save(sess, './foo', global_step=step)
# 0: 1
# 1: 6
# 2: 7
# 3: 3
# 4: 8
# 5: 10
# 6: 12
# 7: 14
# 8: 5
# 9: 17
Then later, if we resume from step 6, we get the same output.
import tensorflow as tf
saver = tf.train.import_meta_graph('./foo-5.meta')
with tf.Session() as sess:
saver.restore(sess, './foo-5')
for step in range(6, 10):
print('{}: {}'.format(step, sess.run('batch:0')))
# 6: 12
# 7: 14
# 8: 5
# 9: 17
|
Hold a candle
Household items Other phrases about:
What's the meaning of the phrase 'Hold a candle'? The expression 'can't hold a candle to' refers to someone who compares badly to an known authority - to be unfit even to hold a subordinate position.
What's the origin of the phrase 'Hold a candle'?
Apprentices used to be expected to hold the candle so that more experienced workmen were able to see what they were doing. Someone unable even to do that would be of low status indeed.
Sir Edward Dering used a similar phrase 'to hold the candle' in his The fower cardinal-vertues of a Carmelite fryar, 1641:
"Though I be not worthy to hold the candle to Aristotle."
'To hold a candle' is first recorded in 1883 in William Norris's No New Thing:
|
/*
* MegaMek - Copyright (C) 2002,2003,2004 Ben Mazur ([email protected])
*
* This program is free software; you can redistribute it and/or modify it under
* the terms of the GNU General Public License as published by the Free Software
* Foundation; either version 2 of the License, or (at your option) any later
* version.
*
* This program is distributed in the hope that it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
* FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
* details.
*/
package megamek.common;
import java.text.NumberFormat;
import java.util.Arrays;
import java.util.Map;
import java.util.Vector;
import megamek.common.options.OptionsConstants;
import megamek.common.preference.PreferenceManager;
import megamek.common.weapons.InfantryAttack;
import megamek.common.weapons.infantry.InfantryWeapon;
/**
* This class represents a squad or point of battle armor equiped infantry,
* sometimes referred to as "Elementals". Much of the behaviour of a battle
* armor unit is identical to that of an infantry platoon, and is rather
* different than that of a Mek or Tank.
*
* @author [email protected] (James A. Damour )
* @version $revision:$
*/
/*
* PLEASE NOTE!!! My programming style is to put constants first in tests so the
* compiler catches my "= for ==" errors.
*/
public class BattleArmor extends Infantry {
/**
*
*/
private static final long serialVersionUID = 4594311535026187825L;
/*
* Infantry have no critical slot limitations. IS squads usually have 4 men,
* Clan points usually have 5. Have a location that represents the entire
* squad.
*/
private static final int[] IS_NUM_OF_SLOTS = { 20, 2, 2, 2, 2, 2, 2 };
private static final String[] IS_LOCATION_ABBRS = { "Squad", "Trooper 1",
"Trooper 2", "Trooper 3", "Trooper 4", "Trooper 5", "Trooper 6" };
private static final String[] IS_LOCATION_NAMES = { "Squad", "Trooper 1",
"Trooper 2", "Trooper 3", "Trooper 4", "Trooper 5", "Trooper 6" };
private static final int[] CLAN_NUM_OF_SLOTS = { 20, 2, 2, 2, 2, 2, 2 };
private static final String[] CLAN_LOCATION_ABBRS = { "Point", "Trooper 1",
"Trooper 2", "Trooper 3", "Trooper 4", "Trooper 5", "Trooper 6" };
private static final String[] CLAN_LOCATION_NAMES = { "Point", "Trooper 1",
"Trooper 2", "Trooper 3", "Trooper 4", "Trooper 5", "Trooper 6" };
public static final int MANIPULATOR_NONE = 0;
public static final int MANIPULATOR_ARMORED_GLOVE = 1;
public static final int MANIPULATOR_BASIC = 2;
public static final int MANIPULATOR_BASIC_MINE_CLEARANCE = 3;
public static final int MANIPULATOR_BATTLE = 4;
public static final int MANIPULATOR_BATTLE_MAGNET = 5;
public static final int MANIPULATOR_BATTLE_VIBRO = 6;
public static final int MANIPULATOR_HEAVY_BATTLE = 7;
public static final int MANIPULATOR_HEAVY_BATTLE_MAGNET = 8;
public static final int MANIPULATOR_HEAVY_BATTLE_VIBRO = 9;
public static final int MANIPULATOR_SALVAGE_ARM = 10;
public static final int MANIPULATOR_CARGO_LIFTER = 11;
public static final int MANIPULATOR_INDUSTRIAL_DRILL = 12;
/**
* A list of the internal names for the different manipulator types.
* The indices in this collection correspond to the MANIPULATOR defines
* in <code>BattleArmor</code>. These names should match the internal
* name for the manipulator's MiscType entry.
*/
public static final String[] MANIPULATOR_TYPE_STRINGS = { "None",
"BAArmoredGlove", "BABasicManipulator",
"BABasicManipulatorMineClearance", "BABattleClaw",
"BABattleClawMagnets", "BABattleClawVibro",
"BAHeavyBattleClaw", "BAHeavyBattleClawMagnets",
"BAHeavyBattleClawVibro",
"BASalvageArm", "BACargoLifter", "BAIndustrialDrill" };
/**
* A list of the display names for the different manipulator types.
* The indices in this collection correspond to the MANIPULATOR defines
* in <code>BattleArmor</code>. These names should match the
* name for the manipulator's MiscType entry.
*/
public static final String[] MANIPULATOR_NAME_STRINGS = { "None",
"BA Manipulators [Armored Gloves]", "BA Manipulators [Manipulator (Basic)]",
"BA Manipulator Adaptation [Mine Clearance Equipment]", "BA Manipulators [Battle Claw]",
"BA Manipulator Adaptation [Magnetic Battle Claw]", "BA Manipulator Adaptation [Vibro-Claw]",
"BA Manipulators [Heavy Battle Claw]", "BA Manipulator Adaptation [Heavy Magnetic Battle Claw]",
"BA Manipulator Adaptation [Heavy Vibro-Claw]", "BA Manipulators [Salvage Arm]", "BA Manipulators [Cargo Lifter]",
"BA Manipulators [Industrial Drill]" };
public static final int CHASSIS_TYPE_BIPED = 0;
public static final int CHASSIS_TYPE_QUAD = 1;
/**
* The number of men alive in this unit at the beginning of the phase,
* before it begins to take damage.
*/
private int troopersShooting = 0;
/**
* the number of troopers of this squad, dead or alive
*/
private int troopers = -1;
/**
* The cost of this unit. This value should be set when the unit's file is
* read.
*/
protected int myCost = -1;
/**
* This unit's weight class
*/
private int weightClass = -1;
/**
* this unit's chassis type, should be BattleArmor.CHASSIS_TYPE_BIPED or
* BattleArmor.CHASSIS_TYPE_QUAD
*/
private int chassisType = -1;
/**
* Flag that is <code>true</code> when this object's constructor has
* completed.
*/
private boolean isInitialized = false;
/**
* Flag that is <code>true</code> when this unit is equipped with stealth.
*/
private boolean isStealthy = false;
/**
* Flag that is <code>true</code> when this unit is equipped with mimetic
* armor.
*/
private boolean isMimetic = false;
/**
* Flag that is <code>true</code> when this unit is equipped with a camo
* system.
*/
private boolean hasCamoSystem = false;
/**
* Modifiers to <code>ToHitData</code> for stealth.
*/
private int shortStealthMod = 0;
private int mediumStealthMod = 0;
private int longStealthMod = 0;
private String stealthName = null;
private String camoName = null;
// Public and Protected constants, constructors, and methods.
/**
* Internal name of the Inner disposable SRM2 ammo pack.
*/
public static final String DISPOSABLE_SRM2_AMMO = "BA-SRM2 (one shot) Ammo";
/**
* Internal name of the disposable NARC ammo pack.
*/
public static final String DISPOSABLE_NARC_AMMO = "BA-Compact Narc Ammo";
/**
* The internal name for the Mine Launcher weapon.
*/
public static final String MINE_LAUNCHER = "BAMineLauncher";
/**
* The internal name for advanced.
*/
public static final String ADVANCED_ARMOR = EquipmentType
.getArmorTypeName(EquipmentType.T_ARMOR_BA_STANDARD_ADVANCED);
/**
* The internal name for standard Prototype.
*/
public static final String STANDARD_PROTOTYPE = EquipmentType
.getArmorTypeName(EquipmentType.T_ARMOR_BA_STANDARD_PROTOTYPE);
/**
* The internal name for stealth Prototype.
*/
public static final String STEALTH_PROTOTYPE = EquipmentType
.getArmorTypeName(EquipmentType.T_ARMOR_BA_STEALTH_PROTOTYPE);
/**
* The internal name for basic Stealth armor.
*/
public static final String BASIC_STEALTH_ARMOR = EquipmentType
.getArmorTypeName(EquipmentType.T_ARMOR_BA_STEALTH_BASIC);
/**
* The internal name for standard Stealth armor.
*/
public static final String STANDARD_STEALTH_ARMOR = EquipmentType
.getArmorTypeName(EquipmentType.T_ARMOR_BA_STEALTH);
/**
* The internal name for improved Stealth armor.
*/
public static final String IMPROVED_STEALTH_ARMOR = EquipmentType
.getArmorTypeName(EquipmentType.T_ARMOR_BA_STEALTH_IMP);
/**
* The internal name for Mimetic armor.
*/
public static final String MIMETIC_ARMOR = EquipmentType
.getArmorTypeName(EquipmentType.T_ARMOR_BA_MIMETIC);
/**
* The internal name for fire-resistant armor.
*/
public static final String FIRE_RESISTANT = EquipmentType
.getArmorTypeName(EquipmentType.T_ARMOR_BA_FIRE_RESIST);
/**
* The internal name for Simple Camo equipment.
*/
public static final String CAMO_SYSTEM = "Camo System";
/**
* The internal name for Single-Hex ECM equipment.
*/
public static final String SINGLE_HEX_ECM = "Single-Hex ECM";
/**
* /** The maximum number of men in a battle armor squad.
*/
public static final int BA_MAX_MEN = 6;
/**
* The location for infantry equipment.
*/
public static final int LOC_SQUAD = 0;
public static final int LOC_TROOPER_1 = 1;
public static final int LOC_TROOPER_2 = 2;
public static final int LOC_TROOPER_3 = 3;
public static final int LOC_TROOPER_4 = 4;
public static final int LOC_TROOPER_5 = 5;
public static final int LOC_TROOPER_6 = 6;
/**
* The location for mounted equipment on BA
*/
public static final int MOUNT_LOC_NONE = -1;
public static final int MOUNT_LOC_BODY = 0;
public static final int MOUNT_LOC_RARM = 1;
public static final int MOUNT_LOC_LARM = 2;
public static final int MOUNT_LOC_TURRET = 3;
public static final String[] MOUNT_LOC_NAMES = { "Body", "Right Arm",
"Left Arm", "Turret" };
/**
* How many mount locations are possible?
*/
public static final int MOUNT_NUM_LOCS = 4;
// Quad BA can add critical space by adding a turret mount.
private int turretSize = 0;
private boolean modularTurret = false;
private boolean exoskeleton = false;
/**
* Clan industrial exoskeletons can opt to not use Harjel, to allow them to
* use IS chassis weight; this flag indicates whether or not this is the
* case.
*/
private boolean clanExoWithoutHarjel = false;
@Override
public String[] getLocationAbbrs() {
if (!isInitialized || isClan()) {
return CLAN_LOCATION_ABBRS;
}
return IS_LOCATION_ABBRS;
}
public String[] getBaMountLocAbbr() {
return MOUNT_LOC_NAMES;
}
public static String getBaMountLocAbbr(int loc) {
if (loc == MOUNT_LOC_NONE) {
return "None";
}
return MOUNT_LOC_NAMES[loc];
}
@Override
public String[] getLocationNames() {
if (!isInitialized || isClan()) {
return CLAN_LOCATION_NAMES;
}
return IS_LOCATION_NAMES;
}
/**
* Returns the number of Troopers in the BattleArmor squad, since locations
* for BattleArmor correspond to the different suits instead of the actual
* mount locations for equipment.
*/
@Override
public int locations() {
int retVal = getTroopers();
if (retVal == 0) {
// Return one more than the maximum number of men in the unit.
if (!isInitialized) {
retVal = 6 + 1;
} else if (isClan()) {
retVal = 5 + 1;
} else {
retVal = 4 + 1;
}
} else {
retVal++;
}
return retVal;
}
/**
* Generate a new, blank, battle armor unit. Hopefully, we'll be loaded from
* somewhere.
*/
public BattleArmor() {
// Instantiate the superclass.
super();
setArmorType(EquipmentType.T_ARMOR_BA_STANDARD);
// BA are always one squad
squadn = 1;
// All Battle Armor squads are Clan until specified otherwise.
setTechLevel(TechConstants.T_CLAN_TW);
// Construction complete.
isInitialized = true;
}
@Override
public int getUnitType() {
return UnitType.BATTLE_ARMOR;
}
protected static final TechAdvancement[] TA_BATTLEARMOR = {
new TechAdvancement(TECH_BASE_ALL).setISAdvancement(2710, DATE_NONE, 3058, 2766, 2905)
.setClanAdvancement(2710, DATE_NONE, 3058).setPrototypeFactions(F_TH)
.setReintroductionFactions(F_CS).setTechRating(RATING_D)
.setAvailability(RATING_F, RATING_X, RATING_E, RATING_D)
.setStaticTechLevel(SimpleTechLevel.STANDARD), //ultralight
new TechAdvancement(TECH_BASE_ALL).setISAdvancement(DATE_NONE, 3050, 3050)
.setClanAdvancement(2865, 2870, 2900).setPrototypeFactions(F_CWF)
.setProductionFactions(F_CIH, F_FS, F_LC).setClanApproximate(true, false, false)
.setTechRating(RATING_E)
.setAvailability(RATING_X, RATING_F, RATING_E, RATING_D)
.setStaticTechLevel(SimpleTechLevel.STANDARD), //light
new TechAdvancement(TECH_BASE_ALL).setISAdvancement(2864, 3052, 3052)
.setClanAdvancement(2840, 2868, 2875)
.setClanApproximate(true, false, false).setPrototypeFactions(F_CGS)
.setProductionFactions(F_CWF, F_FS, F_LC, F_CS).setTechRating(RATING_E)
.setAvailability(RATING_X, RATING_D, RATING_D, RATING_D)
.setStaticTechLevel(SimpleTechLevel.STANDARD), //medium
new TechAdvancement(TECH_BASE_ALL).setISAdvancement(DATE_NONE, 3050, 3058)
.setClanAdvancement(2867, 2875, 3058)
.setClanApproximate(true, false, false).setPrototypeFactions(F_CWF)
.setProductionFactions(F_CHH, F_FS, F_LC).setTechRating(RATING_E)
.setAvailability(RATING_X, RATING_F, RATING_E, RATING_D)
.setStaticTechLevel(SimpleTechLevel.STANDARD), // heavy
new TechAdvancement(TECH_BASE_ALL).setISAdvancement(DATE_NONE, 3058, 3060)
.setClanAdvancement(2870, 2877, 3060)
.setClanApproximate(true, false, false).setPrototypeFactions(F_CNC)
.setProductionFactions(F_CGB, F_DC).setTechRating(RATING_E)
.setAvailability(RATING_X, RATING_F, RATING_E, RATING_D)
.setStaticTechLevel(SimpleTechLevel.STANDARD) // assault
};
public static TechAdvancement getConstructionTechAdvancement(int weightClass) {
return new TechAdvancement(TA_BATTLEARMOR[weightClass]);
}
@Override
public TechAdvancement getConstructionTechAdvancement() {
int index = getWeightClass();
if (index < 0 || index >= TA_BATTLEARMOR.length) {
index = EntityWeightClass.WEIGHT_MEDIUM;
}
return TA_BATTLEARMOR[index];
}
/**
* Returns this entity's original jumping mp.
*/
@Override
public int getOriginalJumpMP() {
return jumpMP;
}
/**
* Returns this entity's walking mp, factored for extreme temperatures and
* gravity.
*/
@Override
public int getWalkMP(boolean gravity, boolean ignoreheat,
boolean ignoremodulararmor) {
return getWalkMP(gravity, ignoreheat, ignoremodulararmor, false, false);
}
public int getWalkMP(boolean gravity, boolean ignoreheat,
boolean ignoremodulararmor, boolean ignoreDWP,
boolean ignoreMyomerBooster) {
int j = getOriginalWalkMP();
if (hasMyomerBooster()) {
if (!ignoreMyomerBooster) {
if (getWeightClass() >= EntityWeightClass.WEIGHT_HEAVY) {
j++;
} else {
j += 2;
}
}
} else if (hasWorkingMisc(MiscType.F_MECHANICAL_JUMP_BOOSTER)) {
// mechanical jump booster gives an extra MP
j++;
}
if (hasDWP() && !ignoreDWP) {
if (getWeightClass() == EntityWeightClass.WEIGHT_MEDIUM) {
j -= 3;
} else if (getWeightClass() >= EntityWeightClass.WEIGHT_HEAVY) {
j -= 2;
}
if (j == 0) {
j++;
}
}
if (null != game) {
int weatherMod = game.getPlanetaryConditions()
.getMovementMods(this);
if (weatherMod != 0) {
j = Math.max(j + weatherMod, 0);
}
}
if (gravity) {
j = applyGravityEffectsOnMP(j);
}
return j;
}
@Override
public int getRunMP(boolean gravity, boolean ignoreheat,
boolean ignoremodulararmor) {
boolean fastMove = (game != null) &&
game.getOptions().booleanOption(OptionsConstants.ADVGRNDMOV_TACOPS_FAST_INFANTRY_MOVE);
if(fastMove) {
return getWalkMP(gravity, ignoreheat, ignoremodulararmor, false,
false) + 1;
}
return getWalkMP(gravity, ignoreheat, ignoremodulararmor, false, false);
}
/**
* does this ba mount a myomer booster?
*
* @return
*/
public boolean hasMyomerBooster() {
for (Mounted mEquip : getMisc()) {
MiscType mtype = (MiscType) mEquip.getType();
if (mtype.hasFlag(MiscType.F_MASC) && !mEquip.isInoperable()) {
return true;
}
}
return false;
}
/*
* (non-Javadoc)
*
* @see megamek.common.Infantry#getJumpMP(boolean)
*/
@Override
public int getJumpMP(boolean gravity) {
return getJumpMP(gravity, false, false);
}
/**
* get this BA's jump MP, possibly ignoring gravity and burden
*
* @param gravity
* @param ignoreBurden
* @return
*/
public int getJumpMP(boolean gravity, boolean ignoreBurden,
boolean ignoreDWP) {
if (isBurdened() && !ignoreBurden) {
return 0;
}
if (hasDWP() && !ignoreDWP) {
return 0;
}
if (null != game) {
int windCond = game.getPlanetaryConditions().getWindStrength();
if (windCond >= PlanetaryConditions.WI_STORM) {
return 0;
}
}
int mp = 0;
if (getMovementMode() != EntityMovementMode.INF_UMU) {
mp = getOriginalJumpMP();
}
// if we have no normal jump jets, we get 1 jump MP from mechanical jump
// boosters, if we have them.
if ((mp == 0) && hasWorkingMisc(MiscType.F_MECHANICAL_JUMP_BOOSTER)) {
mp++;
}
// partial wing gives extra MP in atmosphere
if ((mp > 0)
&& hasWorkingMisc(MiscType.F_PARTIAL_WING)
&& ((game == null) || !game.getPlanetaryConditions().isVacuum())) {
mp++;
}
if ((mp > 0) && hasWorkingMisc(MiscType.F_JUMP_BOOSTER)) {
// jump booster gives an extra MP
mp++;
}
if (gravity) {
mp = applyGravityEffectsOnMP(mp);
}
return mp;
}
/**
* Returns the name of the type of movement used. This is Infantry-specific.
*/
@Override
public String getMovementString(EntityMovementType mtype) {
switch (mtype) {
case MOVE_NONE:
return "None";
case MOVE_WALK:
case MOVE_RUN:
return "Walked";
case MOVE_VTOL_WALK:
case MOVE_VTOL_RUN:
return "Flew";
case MOVE_JUMP:
return "Jumped";
default:
return "Unknown!";
}
}
/**
* Returns the abbreviation of the type of movement used. This is
* Infantry-specific.
*/
@Override
public String getMovementAbbr(EntityMovementType mtype) {
switch (mtype) {
case MOVE_NONE:
return "N";
case MOVE_WALK:
return "W";
case MOVE_RUN:
return "R";
case MOVE_JUMP:
return "J";
case MOVE_VTOL_WALK:
case MOVE_VTOL_RUN:
return "F";
default:
return "?";
}
}
/**
* Battle Armor units can only get hit in undestroyed troopers.
*/
@Override
public HitData rollHitLocation(int table, int side, int aimedLocation,
int aimingMode, int cover) {
// If this squad was killed, target trooper 1 (just because).
if (isDoomed()) {
return new HitData(1);
}
if ((aimedLocation != LOC_NONE)
&& (aimingMode != IAimingModes.AIM_MODE_NONE)) {
int roll = Compute.d6(2);
if ((5 < roll) && (roll < 9)) {
return new HitData(aimedLocation, side == ToHitData.SIDE_REAR,
true);
}
}
// Pick a random number between 1 and 6.
int loc = Compute.d6();
// Pick a new random number if that trooper is dead or never existed.
// Remember that there's one more location than the number of troopers.
// In http://forums.classicbattletech.com/index.php/topic,43203.0.html,
// "previously destroyed includes the current phase" for rolling hits on
// a squad,
// modifying previous ruling in the AskThePM FAQ.
while ((loc >= locations())
|| (IArmorState.ARMOR_NA == this.getInternal(loc))
|| (IArmorState.ARMOR_DESTROYED == this.getInternal(loc))
|| ((IArmorState.ARMOR_DOOMED == this.getInternal(loc)) && !isDoomed())) {
loc = Compute.d6();
}
int critLocation = Compute.d6();
// TacOps p. 108 Trooper takes a crit if a second roll is the same
// location as the first.
if (game.getOptions().booleanOption(OptionsConstants.ADVCOMBAT_TACOPS_BA_CRITICALS)
&& (loc == critLocation)) {
return new HitData(loc, false, HitData.EFFECT_CRITICAL);
}
// Hit that trooper.
return new HitData(loc);
}
@Override
public HitData rollHitLocation(int table, int side) {
return rollHitLocation(table, side, LOC_NONE,
IAimingModes.AIM_MODE_NONE, LosEffects.COVER_NONE);
}
/**
* For level 3 rules, each trooper occupies a specific location
* precondition: hit is a location covered by BA
*/
@Override
public HitData getTrooperAtLocation(HitData hit, Entity transport) {
if (transport instanceof Mech) {
int loc = 99;
switch (hit.getLocation()) {
case Mech.LOC_RT:
if (hit.isRear()) {
loc = 3;
} else {
loc = 1;
}
break;
case Mech.LOC_LT:
if (hit.isRear()) {
loc = 4;
} else {
loc = 2;
}
break;
case Mech.LOC_CT:
if (hit.isRear()) {
loc = 5;
} else {
loc = 6;
}
break;
}
if (loc < locations()) {
return new HitData(loc);
}
} else if (transport instanceof Tank) {
int loc = 99;
switch (hit.getLocation()) {
case Tank.LOC_RIGHT:
// There are 2 troopers on each location, so pick
// one randomly if both are alive.
if ((getInternal(1) > 0) && (getInternal(2) > 0)) {
loc = Compute.randomInt(2) + 1;
} else if (getInternal(1) > 0) {
loc = 1;
} else {
loc = 2;
}
break;
case Tank.LOC_LEFT:
if ((getInternal(3) > 0) && (getInternal(4) > 0)) {
loc = Compute.randomInt(2) + 3;
} else if (getInternal(3) > 0) {
loc = 3;
} else {
loc = 4;
}
break;
case Tank.LOC_REAR:
//Troopers 5 and 6 only exist when you have Clan and CS/WoB units, so we need
//to ensure the array is large enough before checking their status.
if (locations() >= 7 && ((getInternal(5) > 0) && (getInternal(6) > 0))) {
//If we have a live trooper 5 and 6, randomize who gets hit
loc = Compute.randomInt(2) + 5;
} else if ((locations() >= 7) && (getInternal(5) <= 0)) {
//If trooper 5 is dead, hit trooper 6
loc = 6;
} else {
loc = 5;
}
break;
}
//If we get here with an invalid loc for this squad, this next test should fail
if (loc < locations()) {
return new HitData(loc);
}
}
// otherwise roll a random location
return rollHitLocation(ToHitData.HIT_NORMAL, ToHitData.SIDE_FRONT);
}
/**
* Battle Armor units don't transfer damage.
*/
@Override
public HitData getTransferLocation(HitData hit) {
// If any trooper lives, the unit isn't destroyed.
for (int loop = 1; loop < locations(); loop++) {
if (0 < this.getInternal(loop)) {
return new HitData(Entity.LOC_NONE);
}
}
// No surviving troopers, so we're toast.
return new HitData(Entity.LOC_DESTROYED);
}
/**
* Battle Armor units use default behavior for armor and internals.
*
* @see megamek.common.Infantry#isPlatoon()
*/
@Override
protected boolean isPlatoon() {
return false;
}
/**
* Battle Armor units have no armor on their squad location.
*
* @see megamek.common.Infantry#getArmor(int, boolean )
*/
@Override
public int getArmor(int loc, boolean rear) {
if (BattleArmor.LOC_SQUAD != loc) {
return super.getArmor(loc, rear);
}
return IArmorState.ARMOR_NA;
}
/**
* Battle Armor units have no armor on their squad location.
*
* @see megamek.common.Infantry#getOArmor(int, boolean )
*/
@Override
public int getOArmor(int loc, boolean rear) {
if (BattleArmor.LOC_SQUAD != loc) {
return super.getOArmor(loc, rear);
}
return IArmorState.ARMOR_NA;
}
/**
* Battle Armor units have no internals on their squad location.
*
* @see megamek.common.Infantry#getInternal(int )
*/
@Override
public int getInternal(int loc) {
if (BattleArmor.LOC_SQUAD != loc) {
return super.getInternal(loc);
}
return IArmorState.ARMOR_NA;
}
/**
* Battle Armor units have no internals on their squad location.
*
* @see megamek.common.Infantry#getOInternal(int )
*/
@Override
public int getOInternal(int loc) {
if (BattleArmor.LOC_SQUAD != loc) {
return super.getOInternal(loc);
}
return IArmorState.ARMOR_NA;
}
/**
* Set the troopers in the unit to the appropriate values.
*/
@Override
public void autoSetInternal() {
// No troopers in the squad location.
initializeInternal(IArmorState.ARMOR_NA, LOC_SQUAD);
// Initialize the troopers.
for (int loop = 1; loop < locations(); loop++) {
initializeInternal(1, loop);
}
// Set the initial number of troopers that can shoot
// to one less than the number of locations in the unit.
troopersShooting = locations() - 1;
}
/**
* Set the troopers in the unit to the given values.
*/
public void setInternal(int value) {
// Initialize the troopers.
for (int loop = 1; loop < locations(); loop++) {
initializeInternal(value, loop);
}
}
/**
* Mounts the specified equipment in the specified location.
*/
@Override
public void addEquipment(Mounted mounted, int loc, boolean rearMounted)
throws LocationFullException {
// Implement parent's behavior.
super.addEquipment(mounted, loc, rearMounted);
// Is the item a camo system equipment?
String name = mounted.getType().getInternalName();
if (BattleArmor.CAMO_SYSTEM.equals(name)) {
hasCamoSystem = true;
camoName = name;
}
}
/**
* Battle Armor units have as many critical slots as they need to hold their
* equipment.
*/
@Override
protected int[] getNoOfSlots() {
if(!isInitialized) {
return CLAN_NUM_OF_SLOTS;
}
return Arrays.copyOf(isClan() ? CLAN_NUM_OF_SLOTS : IS_NUM_OF_SLOTS, troopers + 1);
}
/**
* Trooper's equipment dies when they do.
*/
@Override
public boolean hasHittableCriticals(int loc) {
if (LOC_SQUAD == loc) {
return false;
}
return super.hasHittableCriticals(loc);
}
/**
* Calculates the battle value of this platoon.
*/
@Override
public int calculateBattleValue(boolean ignoreC3, boolean ignorePilot) {
if (useManualBV) {
return manualBV;
}
return calculateBattleValue(ignoreC3, ignorePilot, false);
}
/**
* Calculates the battle value of this platoon.
*
* @param ignoreC3
* ignore C3 linkage
* @param ignorePilot
* ignore the skill of the pilot
* @param singleTrooper
* calculate just the BV of a single trooper
* @return the battlevalue
*/
public int calculateBattleValue(boolean ignoreC3, boolean ignorePilot,
boolean singleTrooper) {
if (useManualBV) {
return manualBV;
}
// we do this per trooper, then add up
double squadBV = 0;
for (int i = 1; i < locations(); i++) {
if (this.getInternal(i) <= 0) {
continue;
}
double dBV = 0;
double armorBV = 2.5;
if (isFireResistant() || isReflective() || isReactive()) {
armorBV = 3.5;
}
dBV += (getArmor(i) * armorBV) + 1;
// improved sensors add 1
if (hasImprovedSensors()) {
dBV += 1;
}
// active probes add 1
if (hasActiveProbe()) {
dBV += 1;
}
// ECM adds 1
for (Mounted mounted : getMisc()) {
if (mounted.getType().hasFlag(MiscType.F_ECM)) {
if (mounted.getType().hasFlag(MiscType.F_ANGEL_ECM)) {
dBV += 2;
} else {
dBV += 1;
}
break;
}
}
for (Mounted weapon : getWeaponList()) {
if (weapon.getType().hasFlag(WeaponType.F_AMS)) {
if (weapon.getLocation() == LOC_SQUAD) {
dBV += weapon.getType().getBV(this);
}
else {
// squad support, count at 1/troopercount
dBV += weapon.getType().getBV(this)
/ getTotalOInternal();
}
}
}
int runMP = getWalkMP(false, false, true, true, false);
int umuMP = getActiveUMUCount();
int tmmRan = Compute.getTargetMovementModifier(Math.max(runMP,umuMP), false, false,
game).getValue();
// get jump MP, ignoring burden
int rawJump = getJumpMP(false, true, true);
int tmmJumped = (rawJump > 0) ? Compute.
getTargetMovementModifier(rawJump, true, false, game).
getValue() : 0;
double targetMovementModifier = Math.max(tmmRan, tmmJumped);
double tmmFactor = 1 + (targetMovementModifier / 10) + 0.1;
if (hasCamoSystem) {
tmmFactor += 0.2;
}
if (isStealthy) {
tmmFactor += 0.2;
}
// improved stealth get's an extra 0.1, for 0.3 total
if ((stealthName != null)
&& stealthName.equals(BattleArmor.IMPROVED_STEALTH_ARMOR)) {
tmmFactor += 0.1;
}
if (isMimetic) {
tmmFactor += 0.3;
}
dBV *= tmmFactor;
double oBV = 0;
for (Mounted weapon : getWeaponList()) {
// infantry weapons don't count at all
if (weapon.getType().hasFlag(WeaponType.F_INFANTRY) || weapon.getType().hasFlag(WeaponType.F_AMS)) {
continue;
}
if (weapon.getLocation() == LOC_SQUAD) {
// Squad support, count at 1/troopercount
if (weapon.isSquadSupportWeapon()){
oBV += weapon.getType().getBV(this) / getTotalOInternal();
} else {
oBV += weapon.getType().getBV(this);
}
} else {
oBV += weapon.getType().getBV(this) / getTotalOInternal();
}
}
for (Mounted misc : getMisc()) {
if (misc.getType().hasFlag(MiscType.F_MINE)) {
if (misc.getLocation() == LOC_SQUAD) {
oBV += misc.getType().getBV(this);
} else {
oBV += misc.getType().getBV(this) / getTotalOInternal();
}
}
if (misc.getType().hasFlag(MiscType.F_MAGNETIC_CLAMP)) {
if (misc.getLocation() == LOC_SQUAD) {
oBV += misc.getType().getBV(this);
} else {
oBV += misc.getType().getBV(this) / getTotalOInternal();
}
}
}
for (Mounted ammo : getAmmo()) {
int loc = ammo.getLocation();
// don't count oneshot ammo
if (loc == LOC_NONE) {
continue;
}
if ((loc == LOC_SQUAD) || (loc == i)) {
double ammoBV = ((AmmoType) ammo.getType()).getBABV();
oBV += ammoBV;
}
}
if (canMakeAntiMekAttacks()) {
// all non-missile and non-body mounted direct fire weapons
// counted again
for (Mounted weapon : getWeaponList()) {
// infantry weapons don't count at all
if (weapon.getType().hasFlag(WeaponType.F_INFANTRY) || weapon.getType().hasFlag(WeaponType.F_AMS)) {
continue;
}
if (weapon.getLocation() == LOC_SQUAD) {
if (!weapon.getType().hasFlag(WeaponType.F_MISSILE)
&& !weapon.isBodyMounted()) {
oBV += weapon.getType().getBV(this);
}
} else {
// squad support, count at 1/troopercount
oBV += weapon.getType().getBV(this)
/ getTotalOInternal();
}
}
// magnetic claws and vibro claws counted again
for (Mounted misc : getMisc()) {
if ((misc.getLocation() == LOC_SQUAD)
|| (misc.getLocation() == i)) {
if (misc.getType().hasFlag(MiscType.F_MAGNET_CLAW)
|| misc.getType().hasFlag(MiscType.F_VIBROCLAW)) {
oBV += misc.getType().getBV(this);
}
}
}
}
// getJumpMP won't return UMU MP, so weed need to count that extra
int movement = Math.max(getWalkMP(false, false, true, true, false),
Math.max(getJumpMP(false, true, true), getActiveUMUCount()));
double speedFactor = Math.pow(1 + ((double) (movement - 5) / 10), 1.2);
speedFactor = Math.round(speedFactor * 100) / 100.0;
oBV *= speedFactor;
double soldierBV;
if (useGeometricMeanBV()) {
soldierBV = 2 * Math.sqrt(oBV * dBV);
if (soldierBV == 0) {
soldierBV = oBV + dBV;
}
} else {
soldierBV = oBV + dBV;
}
squadBV += soldierBV;
/*
* if (i == 1) { System.out.println(getChassis()+getModel());
* System.out.println(dBV); System.out.println(oBV);
* System.out.println((oBV+dBV)); }
*/
}
// we have now added all troopers, divide by current strength to then
// multiply by the unit size mod
squadBV /= getShootingStrength();
// we might want to get just the BV of a single trooper
if (singleTrooper) {
return (int) Math.round(squadBV);
}
switch (getShootingStrength()) {
case 1:
break;
case 2:
squadBV *= 2.2;
break;
case 3:
squadBV *= 3.6;
break;
case 4:
squadBV *= 5.2;
break;
case 5:
squadBV *= 7;
break;
case 6:
squadBV *= 9;
break;
}
if (!ignoreC3) {
squadBV += getExtraC3BV((int) Math.round(squadBV));
}
// Adjust BV for crew skills.
double pilotFactor = 1;
if (!ignorePilot) {
pilotFactor = getCrew().getBVSkillMultiplier(isAntiMekTrained(), game);
}
int retVal = (int) Math.round(squadBV * pilotFactor);
return retVal;
}
/**
* Prepare the entity for a new round of action.
*/
@Override
public void newRound(int roundNumber) {
// Perform all base-class behavior.
super.newRound(roundNumber);
// If we're equipped with a Magnetic Mine
// launcher, turn it to single shot mode.
for (Mounted m : getMisc()) {
EquipmentType equip = m.getType();
if (BattleArmor.MINE_LAUNCHER.equals(equip.getInternalName())) {
m.setMode("Single");
}
}
}
/**
* Update the unit to reflect damages taken in this phase.
*/
@Override
public void applyDamage() {
super.applyDamage();
int troopersAlive = 0;
for (int i = 0; i < locations(); i++) {
if (getInternal(i) > 0) {
troopersAlive++;
}
}
troopersShooting = troopersAlive;
}
/**
* Get the number of men in the unit (before damage is applied).
*
* @see megamek.common.Infantry#getShootingStrength
*/
@Override
public int getShootingStrength() {
return troopersShooting;
}
public void setCost(int inC) {
myCost = inC;
}
/**
* Determines if the battle armor unit is burdened with un-jettisoned
* equipment. This can prevent the unit from jumping or using their special
* Anti-Mek attacks.
*
* @return <code>true</code> if the unit hasn't jettisoned its equipment
* yet, <code>false</code> if it has.
*/
public boolean isBurdened() {
// Clan Elemental points are never burdened by equipment.
if (!isClan()) {
// if we have ammo left for a body mounted missile launcher,
// we are burdened
for (Mounted mounted : getAmmo()) {
if (mounted.getUsableShotsLeft() == 0) {
// no shots left, we don't count
continue;
}
// first get the weapon we are linked by
// (so we basically only check the currently loaded
// ammo, but if the weapon has no currently loaded ammo, we're
// fine
Mounted weapon = mounted.getLinkedBy();
if ((weapon != null) && weapon.isBodyMounted()
&& weapon.getType().hasFlag(WeaponType.F_MISSILE)) {
return true;
}
} // Check the next piece of equipment
} // End is-inner-sphere-squad
// Unit isn't burdened.
return false;
}
@Override
public boolean canMakeAntiMekAttacks() {
return !isBurdened() && canDoMechanizedBA()
&& (getWeightClass() < EntityWeightClass.WEIGHT_HEAVY)
&& (getMovementMode() != EntityMovementMode.INF_UMU);
}
/**
* does this BA have an unjettisoned DWP?
*
* @return
*/
public boolean hasDWP() {
for (Mounted mounted : getWeaponList()) {
if (mounted.isDWPMounted()) {
if (mounted.isMissing()) {
continue;
} else if ((mounted.getLinked() != null)
&& (mounted.getLinked().getUsableShotsLeft() > 0)) {
return true;
} else if ((mounted.getLinked() == null)
&& !mounted.isMissing()) {
return true;
}
}
}
return false;
}
/**
* Returns true if this <code>BattleArmor</code> can use a detachable weapon
* pack. A <code>BattleArmor</code> must have 2 or more walking MP and be
* Medium or heavier to mount DWP.
*
* @return
*/
public boolean canMountDWP() {
return (getOriginalWalkMP() >= 2)
&& (getWeightClass() >= EntityWeightClass.WEIGHT_MEDIUM);
}
/**
* Returns the name of the stealth Armor used by the BA. Mostly for
* MegaMekLab Usage.
*
* @return name of the stealth armor.
*/
public String getStealthName() {
return stealthName;
}
public String getCamoName() {
return camoName;
}
/**
* Public interface to the BattleArmors short range stealth modifier
*
* @return shortStealthMod
*/
public int getShortStealthMod() {
return shortStealthMod;
}
/**
* Public interface to the BattleArmors medium range stealth modifier
*
* @return mediumStealthMod
*/
public int getMediumStealthMod() {
return mediumStealthMod;
}
/**
* Public interface to the BattleArmors long range stealth modifier
*
* @return longStealthMod
*/
public int getLongStealthMod() {
return longStealthMod;
}
/**
* Determine if this unit has an active stealth system.
* <p/>
* Sub-classes are encouraged to override this method.
*
* @return <code>true</code> if this unit has a stealth system that is
* currently active, <code>false</code> if there is no stealth
* system or if it is inactive.
*/
@Override
public boolean isStealthActive() {
return (isStealthy || isMimetic || hasCamoSystem);
}
public boolean isMimetic() {
return isMimetic;
}
public boolean hasCamoSystem() {
return hasCamoSystem;
}
public boolean isStealthy() {
return isStealthy;
}
/**
* Determine the stealth modifier for firing at this unit from the given
* range. If the value supplied for <code>range</code> is not one of the
* <code>Entity</code> class range constants, an
* <code>IllegalArgumentException</code> will be thrown.
* <p/>
* Sub-classes are encouraged to override this method.
*
* @param range
* - an <code>int</code> value that must match one of the
* <code>Compute</code> class range constants.
* @param ae
* - the entity making the attack.
* @return a <code>TargetRoll</code> value that contains the stealth
* modifier for the given range.
*/
@Override
public TargetRoll getStealthModifier(int range, Entity ae) {
TargetRoll result = null;
// Note: infantry are immune to stealth, but not camoflage
// or mimetic armor
// Mimetic armor modifier is based upon the number of hexes moved,
// and adds to existing movement modifier (Total Warfare p228):
// 0 hexes moved +3 movement modifier
// 1 hex moved +2 movement modifier
// 2 hexes moved +1 movement modifier
// 3+ hexes moved +0 movement modifier
if (isMimetic && !hasMyomerBooster()) {
int mmod = 3 - delta_distance;
mmod = Math.max(0, mmod);
result = new TargetRoll(mmod, "mimetic armor");
}
// Stealthy units alreay have their to-hit mods defined.
if (isStealthy
&& !((ae instanceof Infantry) && !(ae instanceof BattleArmor))
&& !hasMyomerBooster()) {
switch (range) {
case RangeType.RANGE_MINIMUM:
case RangeType.RANGE_SHORT:
result = new TargetRoll(shortStealthMod, stealthName);
break;
case RangeType.RANGE_MEDIUM:
result = new TargetRoll(mediumStealthMod, stealthName);
break;
case RangeType.RANGE_LONG:
case RangeType.RANGE_EXTREME:
case RangeType.RANGE_LOS:
result = new TargetRoll(longStealthMod, stealthName);
break;
case RangeType.RANGE_OUT:
break;
default:
throw new IllegalArgumentException(
"Unknown range constant: " + range);
}
}
// Simple camo modifier is on top of the movement modifier
// 0 hexes moved +2 movement modifier
// 1 hexes moved +1 movement modifier
// 2+ hexes moved no modifier
// This can also be in addition to any armor except Mimetic!
if (hasCamoSystem && (delta_distance < 2)) {
int mod = Math.max(2 - delta_distance, 0);
if (result == null) {
result = new TargetRoll(mod, "camoflage");
} else {
result.append(new TargetRoll(mod, "camoflage"));
}
}
if (result == null) {
result = new TargetRoll(0, "stealth not active");
}
// Return the result.
return result;
} // End public TargetRoll getStealthModifier( char )
@Override
public double getCost(boolean ignoreAmmo) {
return getCost(ignoreAmmo, true);
}
@Override
public double getAlternateCost() {
return getCost(false, false);
}
public double getCost(boolean ignoreAmmo, boolean includeTrainingAndClan) {
if(getChassis().equals("Longinus Battle Armor")
&& getModel().equals("[Flamer]")
&& !includeTrainingAndClan) {
}
double cost = 0;
switch (weightClass) {
case EntityWeightClass.WEIGHT_MEDIUM:
cost += 100000;
if (getMovementMode() == EntityMovementMode.VTOL) {
cost += getOriginalJumpMP() * 100000;
} else {
cost += getOriginalJumpMP() * 75000;
}
break;
case EntityWeightClass.WEIGHT_HEAVY:
cost += 200000;
if (getMovementMode() == EntityMovementMode.INF_UMU) {
cost += getOriginalJumpMP() * 100000;
} else {
cost += getOriginalJumpMP() * 150000;
}
break;
case EntityWeightClass.WEIGHT_ASSAULT:
cost += 400000;
if (getMovementMode() == EntityMovementMode.INF_UMU) {
cost += getOriginalJumpMP() * 150000;
} else {
cost += getOriginalJumpMP() * 300000;
}
break;
default:
cost += 50000;
cost += 50000 * getOriginalJumpMP();
}
cost += 25000 * (getOriginalWalkMP() - 1);
// damn, manipulators are supposed to be treated as structural costs
// and get multiplied by 1.1 if clan
long manipulatorCost = 0;
for (Mounted mounted : getEquipment()) {
if ((mounted.getType() instanceof MiscType)
&& ((MiscType) mounted.getType()).hasFlag(MiscType.F_BA_MANIPULATOR)) {
long itemCost = (long) mounted.getCost();
manipulatorCost += itemCost;
}
}
cost += manipulatorCost;
double baseArmorCost = 10000;
switch(getArmorType(LOC_TROOPER_1)) {
case EquipmentType.T_ARMOR_BA_STANDARD_ADVANCED:
baseArmorCost = 12500;
break;
case EquipmentType.T_ARMOR_BA_MIMETIC:
case EquipmentType.T_ARMOR_BA_STEALTH:
baseArmorCost = 15000;
break;
case EquipmentType.T_ARMOR_BA_STEALTH_BASIC:
baseArmorCost = 12000;
break;
case EquipmentType.T_ARMOR_BA_STEALTH_IMP:
baseArmorCost = 20000;
break;
case EquipmentType.T_ARMOR_BA_STEALTH_PROTOTYPE:
baseArmorCost = 50000;
break;
case EquipmentType.T_ARMOR_BA_FIRE_RESIST:
case EquipmentType.T_ARMOR_BA_STANDARD_PROTOTYPE:
case EquipmentType.T_ARMOR_BA_STANDARD:
default:
baseArmorCost = 10000;
}
cost += (baseArmorCost * getOArmor(LOC_TROOPER_1));
// training cost and clan mod
if (includeTrainingAndClan) {
if (isClan()) {
cost *= 1.1;
cost += 200000;
} else {
cost += 150000;
}
}
// TODO: we do not track the modular weapons mount for 1000 C-bills in
// the unit files
cost += getWeaponsAndEquipmentCost(ignoreAmmo) - manipulatorCost;
return getSquadSize() * cost;
}
@Override
public boolean hasEiCockpit() {
return true;
}
public void setWeightClass(int inWC) {
switch (inWC) {
case 0:
weightClass = EntityWeightClass.WEIGHT_ULTRA_LIGHT;
break;
case 1:
weightClass = EntityWeightClass.WEIGHT_LIGHT;
break;
case 2:
weightClass = EntityWeightClass.WEIGHT_MEDIUM;
break;
case 3:
weightClass = EntityWeightClass.WEIGHT_HEAVY;
break;
case 4:
weightClass = EntityWeightClass.WEIGHT_ASSAULT;
break;
}
}
public double getTrooperWeight() {
return EntityWeightClass.getClassLimit(getWeightClass(), this);
}
@Override
public int getWeightClass() {
return weightClass;
}
public int getTroopers() {
return troopers;
}
public void setTroopers(int troopers) {
this.troopers = troopers;
// this is also squad size
setSquadSize(troopers);
}
public void setChassisType(int inCT) {
chassisType = inCT;
}
public int getChassisType() {
return chassisType;
}
@Override
public boolean canAssaultDrop() {
return true;
}
@Override
public boolean isNuclearHardened() {
return true;
}
public boolean isTrooperActive(int trooperNum) {
return (getInternal(trooperNum) > 0);
}
public int getNumberActiverTroopers() {
int count = 0;
// Initialize the troopers.
for (int loop = 1; loop < locations(); loop++) {
if (isTrooperActive(loop)) {
count++;
}
}
return count;
}
public int getRandomTrooper() {
Vector<Integer> activeTroops = new Vector<Integer>();
for (int loop = 1; loop < locations(); loop++) {
if (isTrooperActive(loop)) {
activeTroops.add(loop);
}
}
int locInt = Compute.randomInt(activeTroops.size());
return activeTroops.elementAt(locInt);
}
@Override
public boolean loadWeapon(Mounted mounted, Mounted mountedAmmo) {
// BA must carry the ammo in same location as the weapon.
// except for mine launcher mines
// This allows for squad weapons and individual trooper weapons
// such as NARC and the support weapons in TW/TO
AmmoType at = (AmmoType) mountedAmmo.getType();
if (!(at.getAmmoType() == AmmoType.T_MINE)
&& (mounted.getLocation() != mountedAmmo.getLocation())) {
return false;
}
return super.loadWeapon(mounted, mountedAmmo);
}
@Override
public boolean loadWeaponWithSameAmmo(Mounted mounted, Mounted mountedAmmo) {
// BA must carry the ammo in same location as the weapon.
// except for mine launcher mines
// This allows for squad weapons and individual trooper weapons
// such as NARC and the support weapons in TW/TO
AmmoType at = (AmmoType) mountedAmmo.getType();
if (!(at.getAmmoType() == AmmoType.T_MINE)
&& (mounted.getLocation() != mountedAmmo.getLocation())) {
return false;
}
return super.loadWeaponWithSameAmmo(mounted, mountedAmmo);
}
public final String getBLK() {
String newline = "\r\n";
StringBuffer buff = new StringBuffer();
buff.append("<BlockVersion>");
buff.append(newline);
buff.append("1");
buff.append(newline);
buff.append("</BlockVersion>");
buff.append(newline);
buff.append("<UnitType>");
buff.append(newline);
buff.append("BattleArmor");
buff.append(newline);
buff.append("</UnitType>");
buff.append(newline);
buff.append("<name>");
buff.append(newline);
buff.append(getChassis());
buff.append(newline);
buff.append("</name>");
buff.append(newline);
buff.append("<model>");
buff.append(newline);
buff.append(getModel());
buff.append(newline);
buff.append("</model>");
buff.append(newline);
buff.append("<year>");
buff.append(newline);
buff.append(getYear());
buff.append(newline);
buff.append("</year>");
buff.append(newline);
buff.append("<type>");
buff.append(newline);
switch (getTechLevel()) {
case TechConstants.T_INTRO_BOXSET:
buff.append("IS Level 1");
break;
case TechConstants.T_IS_TW_NON_BOX:
buff.append("IS Level 2");
break;
case TechConstants.T_IS_ADVANCED:
buff.append("IS Level 3");
break;
case TechConstants.T_IS_EXPERIMENTAL:
buff.append("IS Level 4");
break;
case TechConstants.T_IS_UNOFFICIAL:
buff.append("IS Level 5");
break;
case TechConstants.T_CLAN_TW:
buff.append("Clan Level 2");
break;
case TechConstants.T_CLAN_ADVANCED:
buff.append("Clan Level 3");
break;
case TechConstants.T_CLAN_EXPERIMENTAL:
buff.append("Clan Level 4");
break;
case TechConstants.T_CLAN_UNOFFICIAL:
buff.append("Clan Level 5");
break;
}
buff.append(newline);
buff.append("</type>");
buff.append(newline);
buff.append("<trooper count>");
buff.append(newline);
buff.append(getWeight());
buff.append(newline);
buff.append("</trooper count>");
buff.append(newline);
buff.append("<weightclass>");
buff.append(newline);
switch (getWeightClass()) {
case EntityWeightClass.WEIGHT_ULTRA_LIGHT:
buff.append("0");
break;
case EntityWeightClass.WEIGHT_LIGHT:
buff.append("1");
break;
case EntityWeightClass.WEIGHT_MEDIUM:
buff.append("2");
break;
case EntityWeightClass.WEIGHT_HEAVY:
buff.append("3");
break;
case EntityWeightClass.WEIGHT_ASSAULT:
buff.append("4");
break;
}
buff.append(newline);
buff.append("</weightclass>");
buff.append(newline);
buff.append("<motion_type>");
buff.append(newline);
switch (getMovementMode()) {
case INF_JUMP:
buff.append("jump");
break;
case INF_LEG:
buff.append("leg");
break;
case VTOL:
buff.append("vtol");
break;
case INF_UMU:
buff.append("submarine");
break;
default:
buff.append("none");
break;
}
buff.append(newline);
buff.append("</motion_type>");
buff.append(newline);
buff.append("<chassis>");
buff.append(newline);
switch (getChassisType()) {
case BattleArmor.CHASSIS_TYPE_BIPED:
buff.append("biped");
break;
case BattleArmor.CHASSIS_TYPE_QUAD:
buff.append("quad");
break;
}
buff.append("</chassis>");
buff.append(newline);
buff.append("<cruiseMP>");
buff.append(newline);
buff.append(getOriginalRunMP());
buff.append(newline);
buff.append("</cruiseMP>");
buff.append(newline);
buff.append("<jumpMP>");
buff.append(newline);
buff.append(getOriginalJumpMP());
buff.append(newline);
buff.append("</jumpMP>");
buff.append(newline);
buff.append("<armor>");
buff.append(newline);
buff.append(getOArmor(LOC_TROOPER_1));
buff.append(newline);
buff.append("</armor>");
buff.append(newline);
if (getTurretCapacity() > 0) {
buff.append("<turret>");
buff.append(newline);
if (hasModularTurretMount()) {
buff.append("Modular:");
} else {
buff.append("Standard:");
}
buff.append(String.valueOf(getTurretCapacity()));
buff.append("</turret>");
buff.append(newline);
}
buff.append("<armor_type>");
buff.append(newline);
buff.append(getArmorType(LOC_SQUAD));
buff.append(newline);
buff.append("</armor_type>");
buff.append(newline);
buff.append("<armor_tech>");
buff.append(newline);
buff.append(getArmorTechLevel(LOC_SQUAD));
buff.append(newline);
buff.append("</armor_tech>");
buff.append(newline);
for (int i = 0; i < locations(); i++) {
boolean found = false;
for (Mounted m : getEquipment()) {
// don't write out swarm and leg attack, those get added
// dynamically
if ((m.getType() instanceof WeaponType)
&& m.getType().hasFlag(WeaponType.F_INFANTRY_ATTACK)) {
continue;
}
if (m.getLocation() == i) {
if (!found) {
found = true;
buff.append("<");
buff.append(getLocationName(i));
buff.append(" equipment>");
buff.append(newline);
}
buff.append(m.getType().getInternalName());
buff.append(newline);
}
}
if (found) {
buff.append("</");
buff.append(getLocationName(i));
buff.append(" equipment>");
buff.append(newline);
}
}
return buff.toString();
}
/*
* (non-Javadoc)
*
* @see megamek.common.Entity#getVibroClaws()
*/
@Override
public int getVibroClaws() {
int claws = 0;
for (Mounted mounted : getMisc()) {
if (mounted.getType().hasFlag(MiscType.F_VIBROCLAW)) {
claws++;
}
}
return claws;
}
/**
* return if this BA has fire resistant armor
*
* @return
*/
public boolean isFireResistant() {
for (Mounted equip : getMisc()) {
if (equip.getType().hasFlag(MiscType.F_FIRE_RESISTANT)) {
return true;
}
}
return false;
}
/**
* return if this BA has laser reflective armor
*
* @return
*/
public boolean isReflective() {
for (Mounted equip : getMisc()) {
if (equip.getType().hasFlag(MiscType.F_REFLECTIVE)) {
return true;
}
}
return false;
}
/**
* return if this BA has reactive armor
* @return
*/
public boolean isReactive() {
for (Mounted equip : getMisc()) {
if (equip.getType().hasFlag(MiscType.F_REACTIVE)) {
return true;
}
}
return false;
}
/**
* return if this BA has improved sensors
*
* @return
*/
public boolean hasImprovedSensors() {
for (Mounted equip : getMisc()) {
if (equip.getType().hasFlag(MiscType.F_BAP)) {
if (equip.getType().getInternalName().equals(Sensor.ISIMPROVED)
|| equip.getType().getInternalName()
.equals(Sensor.CLIMPROVED)) {
return true;
}
}
}
return false;
}
/**
* return if the BA has any kind of active probe
*
* @return
*/
public boolean hasActiveProbe() {
for (Mounted equip : getMisc()) {
if (equip.getType().hasFlag(MiscType.F_BAP)
&& !(equip.getType().getInternalName()
.equals(Sensor.ISIMPROVED) || equip.getType()
.getInternalName().equals(Sensor.CLIMPROVED))) {
return true;
}
}
return false;
}
/*
* (non-Javadoc)
*
* @see megamek.common.Entity#canTransferCriticals(int)
*/
@Override
public boolean canTransferCriticals(int loc) {
// BAs can never transfer crits
return false;
}
/**
* can this BattleArmor ride as Mechanized BA?
*
* @return
*/
public boolean canDoMechanizedBA() {
if (getChassisType() != CHASSIS_TYPE_QUAD) {
if (hasWorkingMisc(MiscType.F_MAGNETIC_CLAMP)) {
return true;
}
int tBasicManipulatorCount = countWorkingMisc(MiscType.F_BASIC_MANIPULATOR);
int tArmoredGloveCount = countWorkingMisc(MiscType.F_ARMORED_GLOVE);
int tBattleClawCount = countWorkingMisc(MiscType.F_BATTLE_CLAW);
switch (getWeightClass()) {
case EntityWeightClass.WEIGHT_ULTRA_LIGHT:
case EntityWeightClass.WEIGHT_LIGHT:
if ((tArmoredGloveCount > 1)
|| (tBasicManipulatorCount > 0)
|| (tBattleClawCount > 0)) {
return true;
}
break;
case EntityWeightClass.WEIGHT_MEDIUM:
if ((tBasicManipulatorCount > 0) || (tBattleClawCount > 0)) {
return true;
}
break;
case EntityWeightClass.WEIGHT_HEAVY:
if ((tBasicManipulatorCount > 0) || (tBattleClawCount > 0)) {
return true;
}
break;
case EntityWeightClass.WEIGHT_ASSAULT:
default:
return false;
}
}
return false;
}
@Override
public double getWeight() {
// If following Total Warfare rules each BA trooper will weigh a ton
// for transport purposes. Following Tactical Operations gives us a
// more realistic weight per trooper
if ((game != null)
&& game.getOptions().booleanOption(OptionsConstants.ADVANCED_TACOPS_BA_WEIGHT)) {
double troopton = troopers;
switch (getWeightClass()) {
case EntityWeightClass.WEIGHT_ULTRA_LIGHT:
troopton = troopers * 0.25;
break;
case EntityWeightClass.WEIGHT_LIGHT:
troopton = troopers * 0.5;
break;
case EntityWeightClass.WEIGHT_MEDIUM:
troopton = troopers * 1.0;
break;
case EntityWeightClass.WEIGHT_HEAVY:
troopton = troopers * 1.5;
break;
case EntityWeightClass.WEIGHT_ASSAULT:
troopton = troopers * 2.0;
break;
default:
troopton = troopers;
}
return troopton;
} else {
return troopers;
}
}
public double getAlternateWeight() {
double troopton;
switch (getWeightClass()) {
case EntityWeightClass.WEIGHT_ULTRA_LIGHT:
troopton = troopers * 0.25;
break;
case EntityWeightClass.WEIGHT_LIGHT:
troopton = troopers * 0.5;
break;
case EntityWeightClass.WEIGHT_MEDIUM:
troopton = troopers * 1.0;
break;
case EntityWeightClass.WEIGHT_HEAVY:
troopton = troopers * 1.5;
break;
case EntityWeightClass.WEIGHT_ASSAULT:
troopton = troopers * 2.0;
break;
default:
troopton = troopers;
}
return troopton;
}
@Override
public int getWeaponArc(int wn) {
return Compute.ARC_360;
}
@Override
public boolean isHardenedArmorDamaged(HitData hit) {
return false;
}
@Override
public boolean hasPatchworkArmor() {
return false;
}
@Override
public void setBattleForceMovement(Map<String,Integer> movement) {
if (hasDWP()) {
movement.put("", getWalkMP());
}
int move = Math.max(getWalkMP(true, false, false, true, false),
getJumpMP(true, true, true));
movement.put(getMovementModeAsBattleForceString(), move);
}
@Override
public void setAlphaStrikeMovement(Map<String,Integer> moves) {
if (getMovementMode().equals(EntityMovementMode.INF_JUMP)) {
moves.put("j", getJumpMP(true, true, true) * 2);
} else if (getMovementMode().equals(EntityMovementMode.INF_UMU)) {
moves.put("s", getActiveUMUCount() * 2);
} else {
moves.put(getMovementModeAsBattleForceString(), getOriginalWalkMP() * 2);
}
}
@Override
public int getBattleForceArmorPoints() {
double armor = 0;
for (int i = 1; i < locations(); i++) {
if (armorType[i] == EquipmentType.T_ARMOR_BA_REACTIVE
|| armorType[i] == EquipmentType.T_ARMOR_BA_REFLECTIVE) {
armor += getArmor(i) * 0.75;
} else {
armor += getArmor(i);
}
}
return (int)Math.round(armor / 30.0);
}
@Override
/**
* Each BA squad has 2 structure points
*/
public int getBattleForceStructurePoints() {
return 2;
}
@Override
public void addBattleForceSpecialAbilities(Map<BattleForceSPA,Integer> specialAbilities) {
super.addBattleForceSpecialAbilities(specialAbilities);
for (Mounted m : getEquipment()) {
if (!(m.getType() instanceof MiscType)) {
continue;
}
if (m.getType().hasFlag(MiscType.F_VISUAL_CAMO)
&& !m.getType().getName().equals(MIMETIC_ARMOR)) {
specialAbilities.put(BattleForceSPA.LMAS, null);
} else if (m.getType().hasFlag(MiscType.F_VEHICLE_MINE_DISPENSER)) {
specialAbilities.merge(BattleForceSPA.MDS, 1, Integer::sum);
} else if (m.getType().hasFlag(MiscType.F_TOOLS)
&& (m.getType().getSubType() & MiscType.S_MINESWEEPER) == MiscType.S_MINESWEEPER) {
specialAbilities.put(BattleForceSPA.MSW, null);
} else if (m.getType().hasFlag(MiscType.F_SPACE_ADAPTATION)) {
specialAbilities.put(BattleForceSPA.SOA, null);
} else if (m.getType().hasFlag(MiscType.F_PARAFOIL)) {
specialAbilities.put(BattleForceSPA.PARA, null);
} else if (m.getType().hasFlag(MiscType.F_MAGNETIC_CLAMP)) {
specialAbilities.put(BattleForceSPA.XMEC, null);
}
}
if (canDoMechanizedBA()) {
specialAbilities.put(BattleForceSPA.MEC, null);
}
switch (getArmorType(0)) {
case EquipmentType.T_ARMOR_BA_MIMETIC:
specialAbilities.put(BattleForceSPA.MAS, null);
break;
case EquipmentType.T_ARMOR_BA_FIRE_RESIST:
specialAbilities.put(BattleForceSPA.FR, null);
break;
}
}
public void setIsExoskeleton(boolean exoskeleton) {
this.exoskeleton = exoskeleton;
}
public boolean isExoskeleton() {
return exoskeleton;
}
@Override
public boolean isCrippled() {
double activeTroopPercent = (double) getNumberActiverTroopers()
/ getSquadSize();
if (activeTroopPercent < 0.5) {
if (PreferenceManager.getClientPreferences().debugOutputOn()) {
System.out.println(getDisplayName()
+ " CRIPPLED: only "
+ NumberFormat.getPercentInstance().format(
activeTroopPercent) + " troops remaining.");
}
return true;
}
return false;
}
@Override
public boolean isDmgHeavy() {
return (((double) getNumberActiverTroopers() / getSquadSize()) < 0.67);
}
@Override
public boolean isDmgModerate() {
return (((double) getNumberActiverTroopers() / getSquadSize()) < 0.75);
}
@Override
public boolean isDmgLight() {
return (((double) getNumberActiverTroopers() / getSquadSize()) < 0.9);
}
public int calculateSwarmDamage() {
int damage = 0;
for (Mounted m : getWeaponList()) {
WeaponType wtype;
if (m.getType() instanceof WeaponType) {
wtype = (WeaponType) m.getType();
if (wtype.hasFlag(WeaponType.F_MISSILE)) {
continue;
}
if (m.isBodyMounted()) {
continue;
}
if (wtype instanceof InfantryWeapon) {
continue;
}
if (wtype instanceof InfantryAttack) {
continue;
}
int addToDamage = wtype.getDamage(0);
if (addToDamage < 0) {
continue;
}
// if it's a squad mounted weapon, each trooper hits
if (m.getLocation() == BattleArmor.LOC_SQUAD) {
addToDamage *= getShootingStrength();
}
damage += addToDamage;
}
}
if (hasMyomerBooster()) {
damage += getTroopers() * 2;
}
return damage;
}
public boolean isConventionalInfantry() {
return false;
}
@Override
public long getEntityType(){
return Entity.ETYPE_INFANTRY | Entity.ETYPE_BATTLEARMOR;
}
public int getMaximumJumpMP() {
return getMaximumJumpMP(false);
}
/**
* Returns the maximum jump MP that this BA can have.
*
* @param ignoreEquipment
* If true, bonuses from equipment like partial wing and jump
* booster are ignored. This is important for construction
* purposes, where we shouldn't allow the JSpinner to select
* these values.
* @return
*/
public int getMaximumJumpMP(boolean ignoreEquipment) {
if(chassisType == CHASSIS_TYPE_QUAD) {
return 0;
}
int max = 2;
if(getMovementMode() == EntityMovementMode.INF_UMU) {
if(getWeightClass() <= EntityWeightClass.WEIGHT_LIGHT) {
max = 5;
}
else if(getWeightClass() == EntityWeightClass.WEIGHT_MEDIUM) {
max = 4;
}
else if(getWeightClass() == EntityWeightClass.WEIGHT_HEAVY) {
max = 3;
}
else {
max = 2;
}
}
else if(getMovementMode() == EntityMovementMode.VTOL) {
if(getWeightClass() == EntityWeightClass.WEIGHT_ULTRA_LIGHT) {
max = 7;
}
else if(getWeightClass() == EntityWeightClass.WEIGHT_LIGHT) {
max = 6;
}
else if(getWeightClass() == EntityWeightClass.WEIGHT_MEDIUM) {
max = 5;
}
else {
max = 0;
}
}
else {
if(getWeightClass() < EntityWeightClass.WEIGHT_HEAVY) {
max = 3;
}
}
// Partial wings and jump boosters add 1 jump MP and can increase it
// over the max and they cannot be used together
if (!ignoreEquipment
&& (hasWorkingMisc(MiscType.F_JUMP_BOOSTER)
|| hasWorkingMisc(MiscType.F_PARTIAL_WING))){
max++;
}
return max;
}
public int getMinimumWalkMP() {
if(chassisType == CHASSIS_TYPE_QUAD) {
return 2;
}
return 1;
}
public int getMaximumWalkMP() {
int max = 2;
if(getWeightClass() < EntityWeightClass.WEIGHT_HEAVY) {
max = 3;
}
if(chassisType == CHASSIS_TYPE_QUAD) {
max += 2;
}
// Mechanical jump boosters add 1 MP and can increase it over the max
if (hasWorkingMisc(MiscType.F_MECHANICAL_JUMP_BOOSTER)){
max++;
}
if (hasMyomerBooster()){
switch (getWeightClass()){
case EntityWeightClass.WEIGHT_ULTRA_LIGHT:
case EntityWeightClass.WEIGHT_LIGHT:
case EntityWeightClass.WEIGHT_MEDIUM:
max += 2;
break;
case EntityWeightClass.WEIGHT_HEAVY:
case EntityWeightClass.WEIGHT_ASSAULT:
max++;
break;
}
}
return max;
}
public int getMaximumArmorPoints() {
switch(getWeightClass()) {
case EntityWeightClass.WEIGHT_ULTRA_LIGHT:
return 2;
case EntityWeightClass.WEIGHT_LIGHT:
return 6;
case EntityWeightClass.WEIGHT_MEDIUM:
return 10;
case EntityWeightClass.WEIGHT_HEAVY:
return 14;
case EntityWeightClass.WEIGHT_ASSAULT:
return 18;
default:
return 0;
}
}
@Override
public void setArmorType(int armType) {
for (int i = 0; i < 7; i++) {
armorType[i] = armType;
}
// Set some special state for certain armor types
if(armType == EquipmentType.T_ARMOR_BA_STEALTH_BASIC) {
isStealthy = true;
shortStealthMod = 0;
mediumStealthMod = 1;
longStealthMod = 2;
stealthName = BattleArmor.BASIC_STEALTH_ARMOR;
} else if (armType == EquipmentType.T_ARMOR_BA_STEALTH_PROTOTYPE) {
isStealthy = true;
shortStealthMod = 0;
mediumStealthMod = 1;
longStealthMod = 2;
stealthName = BattleArmor.STEALTH_PROTOTYPE;
} else if (armType == EquipmentType.T_ARMOR_BA_STEALTH) {
isStealthy = true;
shortStealthMod = 1;
mediumStealthMod = 1;
longStealthMod = 2;
stealthName = BattleArmor.STANDARD_STEALTH_ARMOR;
} else if (armType == EquipmentType.T_ARMOR_BA_STEALTH_IMP) {
isStealthy = true;
shortStealthMod = 1;
mediumStealthMod = 2;
longStealthMod = 3;
stealthName = BattleArmor.IMPROVED_STEALTH_ARMOR;
} else if (armType == EquipmentType.T_ARMOR_BA_MIMETIC) {
isMimetic = true;
stealthName = BattleArmor.MIMETIC_ARMOR;
}
}
public int getArmCrits() {
if(getChassisType() == CHASSIS_TYPE_QUAD) {
return 0;
}
switch(getWeightClass()) {
case EntityWeightClass.WEIGHT_ULTRA_LIGHT:
case EntityWeightClass.WEIGHT_LIGHT:
return 2;
case EntityWeightClass.WEIGHT_MEDIUM:
case EntityWeightClass.WEIGHT_HEAVY:
return 3;
default:
return 4;
}
}
public int getBodyCrits() {
if(getChassisType() == CHASSIS_TYPE_QUAD) {
int turret = 0;
if (getTurretCapacity() > 0) {
turret = 1;
if (hasModularTurretMount()) {
turret++;
}
}
switch(getWeightClass()) {
case EntityWeightClass.WEIGHT_ULTRA_LIGHT:
return 0;
case EntityWeightClass.WEIGHT_LIGHT:
return 5 - turret;
case EntityWeightClass.WEIGHT_MEDIUM:
return 7 - turret;
case EntityWeightClass.WEIGHT_HEAVY:
return 9 - turret;
default:
return 11 - turret;
}
} else {
switch(getWeightClass()) {
case EntityWeightClass.WEIGHT_ULTRA_LIGHT:
return 2;
case EntityWeightClass.WEIGHT_LIGHT:
case EntityWeightClass.WEIGHT_MEDIUM:
return 4;
default:
return 6;
}
}
}
public int getTurretCapacity() {
return turretSize;
}
public void setTurretSize(int capacity) {
turretSize = capacity;
}
public boolean hasModularTurretMount() {
return modularTurret;
}
public void setModularTurret(boolean modular) {
modularTurret = modular;
}
public int getTotalCrits() {
return (getArmCrits() * 2) + getBodyCrits() + getTurretCapacity();
}
public int getNumCrits(int loc){
if (loc == MOUNT_LOC_BODY){
return getBodyCrits();
} else if ((loc == MOUNT_LOC_LARM) || (loc == MOUNT_LOC_RARM)){
return getArmCrits();
} else if (loc == MOUNT_LOC_TURRET) {
return getTurretCapacity();
} else {
return 0;
}
}
/**
* Returns the number of allowed anti-mech weapons the supplied location
* can mount. The body can mount a set number of anti-mech weapons and a
* set number of anti-personnel, however for the arms can mount 2 AP or
* 1 AP and 1 AM.
*
* @param loc
* @return
*/
public int getNumAllowedAntiMechWeapons(int loc){
if ((loc == MOUNT_LOC_LARM) || (loc == MOUNT_LOC_RARM)){
return 1;
} else if ((loc == MOUNT_LOC_BODY) || (loc == MOUNT_LOC_TURRET)) {
if (getChassisType() == CHASSIS_TYPE_QUAD){
return 4;
} else {
return 2;
}
} else {
return 0;
}
}
/**
* Returns the number of allowed anti-personnel weapons the location can
* mount. The body can mount a set number of anti-mech weapons and a set
* number of anti-personnel, however the arms can mount 2 AP or 1 AP and 1
* AM.
*
* @param loc
* @return
*/
public int getNumAllowedAntiPersonnelWeapons(int loc, int trooper) {
if ((loc == MOUNT_LOC_LARM) || (loc == MOUNT_LOC_RARM)) {
boolean hasAntiMech = false;
for (Mounted m : getWeaponList()) {
if (!m.getType().hasFlag(WeaponType.F_INFANTRY)
&& (m.getBaMountLoc() == loc)
&& ((m.getLocation() == LOC_SQUAD) || (m.getLocation() == trooper))) {
hasAntiMech = true;
}
}
if (hasAntiMech) {
return 1;
} else {
return 2;
}
} else if (loc == MOUNT_LOC_BODY) {
if (getChassisType() == CHASSIS_TYPE_QUAD) {
return 4;
} else {
return 2;
}
} else {
return 0;
}
}
@Override
public boolean doomedInExtremeTemp() {
return false;
}
@Override
public boolean doomedInVacuum() {
return false;
}
/**
* Convenience method for determining if the BA has magnetic clamps.
*
* @return true if the unit has at least one magnetic clamp, else false
*/
public boolean hasMagneticClamps() {
return countWorkingMisc(MiscType.F_MAGNETIC_CLAMP) > 0;
}
/**
* Returns the <code>EquipmentType</code> internal name for the manipulator
* mounted in the left arm of this <code>BattleArmor</code> squad.
*
* @return
*/
public String getLeftManipulatorName(){
Mounted m = getLeftManipulator();
if (m == null){
return MANIPULATOR_TYPE_STRINGS[MANIPULATOR_NONE];
} else {
return m.getType().getInternalName();
}
}
/**
* Returns the <code>EquipmentType</code> internal name for the manipulator
* mounted in the right arm of this <code>BattleArmor</code> squad.
*
* @return
*/
public String getRightManipulatorName(){
Mounted m = getRightManipulator();
if (m == null){
return MANIPULATOR_TYPE_STRINGS[MANIPULATOR_NONE];
} else {
return m.getType().getInternalName();
}
}
/**
* Returns the <code>Mounted</code> for the manipulator
* mounted in the left arm of this <code>BattleArmor</code> squad.
*
* @return
*/
public Mounted getLeftManipulator(){
for (Mounted m : getMisc()){
if (m.getType().hasFlag(MiscType.F_BA_MANIPULATOR)
&& (m.getBaMountLoc() == MOUNT_LOC_LARM)){
return m;
}
}
return null;
}
/**
* Returns the <code>Mounted</code> for the manipulator
* mounted in the right arm of this <code>BattleArmor</code> squad.
*
* @return
*/
public Mounted getRightManipulator(){
for (Mounted m : getMisc()){
if (m.getType().hasFlag(MiscType.F_BA_MANIPULATOR)
&& (m.getBaMountLoc() == MOUNT_LOC_RARM)){
return m;
}
}
return null;
}
public boolean isClanExoWithoutHarjel() {
return clanExoWithoutHarjel;
}
public void setClanExoWithoutHarjel(boolean clanExoWithoutHarjel) {
this.clanExoWithoutHarjel = clanExoWithoutHarjel;
}
@Override
public String getLocationDamage(int loc) {
String toReturn = "";
if(getInternal(loc)<0) {
return toReturn;
}
boolean first = true;
for(Mounted m : getEquipment()) {
if(m.isMissingForTrooper(loc)) {
if (!first) {
toReturn += ", ";
}
toReturn += m.getName();
first = false;
}
}
return toReturn;
}
/**
* Used to determine the draw priority of different Entity subclasses.
* This allows different unit types to always be draw above/below other
* types.
*
* @return
*/
public int getSpriteDrawPriority() {
return 2;
}
} // End public class BattleArmor extends Infantry implements Serializable
|
Links
Images
Classifications
G11C14/00—Digital stores characterised by arrangements of cells having volatile and non-volatile storage properties for back-up when the power is down
Abstract
A non-volatile semiconductor memory device includes a memory cell array capable of storing data in volatile and non-volatile states, and a determination circuit for determining whether a row address in a current cycle is the same as that in a previous cycle. When the row address in the current cycle is different from that in the previous cycle, a store operation is performed, where volatile data is stored in the memory cell as non-volatile data; thereafter a recall operation is performed where non-volatile data stored in the memory cell is transformed into volatile data; the resulting volatile data is stored in the memory cell and a sense amplifier, or at least one of the sense amplifier and a latch circuit. When the row address in the current cycle is the same as that in the previous cycle, a read or write operation is performed with respect to volatile data stored in the memory cell, the sense amplifier or latch circuit, but non-volatile data is not read from or written to the memory cell.
Description
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a non-volatile memory device including a memory cell having volatile and non-volatile data storage capacity. More particularly, the present invention relates to a Non-Volatile Dynamic Random Access Memory (hereinafter referred to as “NVDRAM”) including a memory cell having a combination of a Dynamic Random Access Memory (hereinafter referred to as “DRAM”) as a volatile data storage means and a Electrically Erasable and Programmable Read Only Memory (hereinafter referred to as “EEPROM”) as a non-volatile data storage means, or a memory cell having volatile and non-volatile data storage capacity using a ferroelectric material. Further, the present invention relates to a system Large-Scale Integrated circuit (hereinafter referred to as “LSI”) such as a system control circuit or system solution chip; a system Center Processing Unit (hereinafter referred to as “CPU”); a system Multi Processing Unit (hereinafter referred to as “MPU”); and the like.
2. Description of the Related Art
Conventional NVDRAMs are typically categorized into two types. One type includes a combination of a DRAM and a EEPROM. The other type employs a ferroelectric material.
In a normal volatile (DRAM) operation, the DRAM and EPPROM combination type NVDRAM uses its DRAM for data access. In a store operation, immediately before the NVDRAM is turned off or at any time desired, data held in the DRAM is transferred to the EEPROM. In the recall operation, the data stored in the non-volatile state is transferred from the EEPROM to the DRAM.
As a result, a high-speed data access can be achieved by the DRAM in the normal DRAM operation while the data can be saved safely in the EEPROM in the absence of supplied power. The advantages of the combination type NVDRAM are that data can be retained separately in the DRAM and the EEPROM; and that there is no limitation to the number of times the recall operation can be repeated.
Article (1) describes a ferroelectric type NVDRAM as including memory cells each having two transistors and two capacitors (hereinafter this type of NVDRAM will be referred to as a two-transistor/cell NVDRAM). In the two-transistor/cell NVDRAM, a read/write operation is performed in a recall/store mode which involves polarization reversal of a ferroelectric thin film included in the memory cell. However, the two-transistor/cell NVDRAM cannot satisfy the required number of available polarization reversals (1015 to 1016 times) of the ferroelectric thin film in order to achieve 10-year continuous operation with a cycle of 100 nsec. This is because the possible number of the available polarization reversals of the ferroelectric thin film is limited to 1012 to 1013 times at maximum under the preset situation. Article (2) discusses the difference in fatigue of the ferroelectric thin film operated during a volatile (DRAM) operation without the polarization reversal and during a non-volatile (recall/store) operation. It is reported that the fatigue of the ferroelectric thin film is significantly small during the volatile operation compared with during the non-volatile operation. Article (3) describes an example of a non-volatile semiconductor memory device including word lines selected by row addresses and plate lines and sense amplifiers selected by column addresses, realizing low power consumption and a short cycle.
The above-mentioned ferroelectric type NVDRAMs use memory cells each incorporating a capacitor element which has a thin film made of a ferroelectric material with a crystal structure of a perovskite type, e.g., Y1 (the common name of a ferroelectric ceramic which is less fatigued by a rewrite operation (polarization reversal)), PZT, PLZT, PbTiO3, etc. Many kinds of ferroelectric materials have been vigorously developed in order to realize an ideal ferroelectric thin film which is not fatigued by the polarization reversal and which has a large remanent polarization and a large dielectric constant.
When an AC voltage is applied to the capacitor element including such a ferrroelectric thin film, the polarization of the ferroelectric thin film shows a hysteresis characteristic as shown in FIG. 11. As seen from FIG. 11, the polarization state of the ferroelectric material, which is at point A when it is not polarizaed, shifts to point B when a positive electric field is applied to the ferroelectric thin film. The polarization state of the ferroelectric thin film returns only to point C (instead of A) when the electric field is removed, thus resulting in a positive remanent polarization. This remanent polarization vanishes when a negative electric field is applied. By further increasing the intensity of the negative electric field, the polarization of the ferroelectric thin film is reversed so that the polarization state shifts to point D. The polarization state of the ferroelectric thin film returns only to point E when the electric field is again removed, thus resulting in a negative remanent polarization. Thus, by allowing the polarization of the ferroelectric thin film to be reversed so as to attain positive or negative remanent polarization, given data can be stored in a non-volatile manner. Moreover, by simply applying or removing a positive or negative electric field across the capacitor element, it can be ensured that the polarization state of the ferroelectric material of the capacitor element shifts only between points B and C or between points D and E, instead of having polarization reversals. Thus, given data can be stored in a volatile manner, as in the case of a usual DRAM.
The above-described ferroelectric NVDRAM has the advantage that, since the memory cells thereof can be constituted by a smaller number of elements than in the case of a combination type NVDRAM, the cell areas can be reduced, thereby providing for further integration of the device. The ferroelectric NVDRAM has another advantage of low power consumption.
Hereinafter, the structure and operation of an exemplary ferroelectric type NVDRAM, where two-transistor/cell memory cells are used, will be described. The two transistor/cell memory cell is known to be immune to possible variations in the fabrication process.
As shown in FIG. 12, this ferroelectric type NVDRAM includes a plurality of word lines WL and corresponding plate lines PT. The word lines WL are connected to a word line decoder 31. The plate lines PT are connected to a plate decoder 32. The NVDRAM also includes a plurality of pairs of bit lines, bit and bit-bar. Each pair of bit lines, bit and bit-bar, are connected to a sense amplifier 33. In FIG. 12, only one pair of bit lines, bit and bit-bar, and their corresponding sense amplifier 33 are shown.
A memory cell 34 is provided in each portion where one of the word lines WL and its corresponding plate line PT intersect a pair of bit lines, bit and bit-bar. In FIG. 12, only one memory cell 34 is shown. The memory cell 34 includes two capacitor elements C1 and C2 and two selection transistors Q1 and Q2. One of the terminals of the capacitor element C1 is connected to the bit line bit via the transistor Q1. One of the terminals of the capacitor element C2 is connected to the bit line bit-bar via the transistor Q2. The other terminals of the capacitor elements C1, and C2 are connected to the plate line PT. Gates of the transistors Q1 and Q2 are connected to the word line WL.
The above-described ferroelectric type NVDRAM operates in the following manner. The word line decoder 31 and the plate line decoder 32 select, respectively, one of the word lines WL and one of the plate lines PT, in accordance with an address signal input to an address buffer 35. Then, the memory cell 34 is accessed in a mode chosen in accordance with a control signalinput from an NE bar pin (an enable pin for switching to a non-volatile mode), a CE-bar bin (an enable pin for selecting a chip), an OE-bar pin (an output enable pin), a WE-bar pin (a write enable pin) to a control signal input buffer 36. Specifically, the access operation to the memory cell 34 is conducted in one of the following modes: in a DRAM mode (or a volatile mode) for accessing data stored in a volatile state, the NVDRAM is controlled by a DRAM mode timing control circuit 37; in a recall mode for reading and rewriting data stored in a non-volatile state, the NVDRAM is controlled by a recall mode timing control circuit 38; in a store mode for writing data stored in a non-volatile state, the NVDRAM is controlled by a store mode timing control circuit 39. The inputting and outputting of the data to be accessed are conducted through a data I/O interface 40.
Hereinafter, a write operation in the store mode, controlled by the store-mode timing control circuit 39, will be described with reference to FIGS. 11 through 14. As shown in FIG. 13, when data “0” is to be written, for example, a voltage of 0 V and a voltage of 5 V (e.g., supply voltage Vcc) are applied to, respectively, the bit line bit and the bit line bit-bar, and a voltage pulse which varies from 0 V and 5 V and back to 0 V is applied to the plate line PT while keeping the word line WL in an active state. As a result, the polarization state of the ferroelectric material of the capacitor element C1 shifts from point C or point E to point B, and then to point C as shown in FIG. 11. The polarization state of the ferroelectric material of the capacitor element C2 shifts from point D to point E and back to point D as shown in FIG. 11. Accordingly, after application of these voltages is stopped, the ferroelectric materials of the capacitor elements C1 and C2 have the remanent polarizations of points C and E, respectively, so that the data '0 is stored in a nonvolatile manner.
When data “1” is to be written, as shown in FIG. 14, a voltage of 5 V and avoltage of 0 V are applied to, respectively, the bit line bit and the bit line bit-bar, in contrast to the voltages of 0 V and 5 V applied in the above-mentioned case of writing the data “0”. A voltage pulse which varies from 0 V to 5 V and back to 0 V is applied to the plate line PT while keeping the word line WL in an active state, whereby the ferroelectric materials of the capacitor elements C1 and C2 are caused to have the remanent polarization of points E and C, respectively, so that the data “1” is stored in a non-volatile manner. Alternatively, a plurality of voltage pulses which, for example, vary from 0 V to 5 V to 0 V to 5 V to 0 V applied to the plate line PT, whereby obtaining a more reliable store operation.
Next, a read operation of data in the recall mode, controlled by the recall mode timing control circuit 38, will be described with reference to FIGS. 11, 12, and 15. In this operation, as shown in FIG. 15, the pair of bit lines bit and bit-bar are precharged at a potential of 0 V and thereafter are placed in an open state. Then, a voltage which varies from 0 V to 5 V is applied to the plate line PT while keeping the word line WL in an active state. As a result, in the case where data “0” is stored in the memory cell 34, the polarization state of the ferroelectric material of the capacitor element C1 shifts from point C to point B as shown in FIG. 11, while the polarization state of the ferroelectric material of the capacitor element C2 shifts from point E to point B. Thus, the polarization of the ferroelectric material of the capacitor C2 is reversed, so that the potential of the bit line bit-bar, which is connected to the capacitor element C2, becomes higher than that of the bit line bit by hundreds of mV (milivolts). Accordingly, by sensing the difference between the potentials of the bit lines bit and bit-bar by the sense amplifier 33, the data stored in a non-volatile manner can be read out.
However, the polarization states of the capacitor elements C1 and C2 both shift to point B, so that the data which has been stored in a non-volatile manner is lost, resulting in a destructive read-out of the data. Accordingly, in the recall mode, the data which has been recalled can be rewritten in the NVDRAM by storing the data in a non-volatile manner. This can be conducted by, after the pair of bit lines bit and bit-bar are set at 0 V and 5 V by the sense amplifier 33, applying a voltage which varies from 0 V to 5 V and back to 0 V to the plate line PT in the same procedure as that for the store mode. In addition, by maintaining the plate line PT at 0 V after the above procedure, the potentials set for the pair of bit lines bit and bit-bar can be stored in the memory cell 34 as charges, thus realizing storage in a DRAM mode (to be described later). The potential difference between the pair of bit lines bit and bit-bar which is generated in the recall-mode is in proportion to the intensity of the remanent polarization of the ferroelectric material and in inverse proportion to the bit line capacitance. Accordingly, a larger potential difference can be obtained as the remanent polarization increases and as the bit line capacity decreases, thereby facilitating the detection by the sense amplifier 33.
An access operation to the data in the DRAM mode, controlled by the DRAM-mode timing control circuit 37, is achieved by the same procedure as that for a conventional DRAM except that a voltage of 0 V or 5 V is applied to the plate line PT. Thus, the polarization states of the ferroelectric materials of the capacitor element C1 and C2 shift only between point D and point E or between point B and point C (FIG. 11), neither of which results in any polarization reversal. Therefore, the read and write of the data stored in a volatile manner can be conducted by utilizing the charges stored in the capacitor elements C1 and C2 alone, as in the case of a conventional DRAM. Refresh operations of the data can also be conducted.
Although a ferroelectric type NVDRAM using two transistor/cell memory cells was described above, substantially the same principle applies to a ferroelectric NVDRAM including a cell having one transistor and one capacitor (hereinafter this type of NVDRAM will be referred to as a one-transistor/cell NVDRAM), as disclosed in Japanese Laid-Open Publication No. 4-176589, which is suitable for a high integration. In addition, the same principle may be applied to NVDRAMs in which addresses are multiplexed and a row address strobe signal RAS-bar and a column address strobe signal CAS-bar are used, as in a usual case of a DRAM. Multiplexing the addresses has the advantage of reducing the number of terminal pins for the addresses so as to obtain a higher integration of the device.
It is possible to operate the combination type NVDRAM and the ferroelectric type NVDRAM both described above in the store mode for writing data in a non-volatile manner and in the recall mode for reading the data stored in the non-volatile manner. However, the operational speed of the ferroelectric NVDRAM is slower than that of the combination type NVDRAM by the polarization reversal time.
Moreover, in the case of the combination type NVDRAM, available times to rewrite the EEPROM are limited to about one hundred thousand times, so that the EEPROM will soon reach the end of its lifetime if frequent rewrite operations are conducted. For this reason, in the case of the combination type NVDRAM, the EEPROM is prevented from undergoing an excessively large number of rewrite operations by ensuring that only the DRAM is accessed during normal operations. Accordingly, the data stored in the DRAM is saved in the EEPROM only immediately before the NVDRAM device is turned off, and all the data is recalled into the DRAM in a recall mode when the NVDRAM device is turned on the next time.
In the case of the ferroelectric type NVDRAM, the ferroelectric materials included in the capacitor elements C1 and C2 permit only a limited number of polarization reversals so that the recall/store operations are limited to about 1010 to 1012 times. Therefore, the lifetime of the memory cells 34 can expire in a few days by conducting successive accessing at a frequency of about 10 MHz. The number of polarization reversals of the ferroelecric material needs to be about 1015 or more in order to obtain the 10-year continuous non-volatile operation (recall/store operation). To this end, the number of the access operations which result in polarization reversals is minimized by ensuring that access operations in the DRAM mode, which do not result in any polarization reversal, are conducted during normal read/write operations and that the data in the memory cells 34 is stored in a non-volatile manner only before the NVDRAM device is turned off. In the ferroelectric type NVDRAM device, all the data stored in the non-volatile state is recalled to the data stored in the volatile state in the first read operation when the NVDRAM device is turned on the next time. In addition, since the access operation in the DRAM mode does not require varying a voltage level of the plate line PT, the access operation in the DRAM mode has the advantages of shortening the cycle and reducing power consumption.
FIGS. 16 and 17 show structures of conventional NVDRAMs called FRAM (or FeRAM). These FRAMs include row decoders to receive row addresses via row address latches and column decoders to receive column addresses via column address latches. The row and column decodersareprovided around memory cell arrays. A memory cell is selected in accordance with an input address. The memory cell is then accessed in accordance with a mode based on control signals CE-bar, WE-bar, and OE-bar input to a control circuit. Data to be written or read to/from the memory cell is input to a column decoder or output to the outside via I/O0 to I/O7 by an I/O latch and an I/O buffer circuit.
For example, in the case of a parallel FRAM having a small capacity (4 Kbit) shown in FIG. 16, higher address pin terminals A3 to A5 are assigned to a row address while lower address pin terminals A0 to A2 are assigned to a column address. However, in the case of a parallel FRAM having a large capacity (64 Kbit) shown in FIG. 17, higher address pin terminals A3 to A9 and A11 and A12 are assigned to a row address while lower address pin terminals A0 to A2 and A10 are assigned to a column address. In the case of the parallel FRAM having the large capacity (64 Kbit), the numbering of the address pin terminals is not serial.
Japanese Laid-Open Publication No. 7-13859 discloses volatile and non-volatile semiconductor memory devices having different capacities and widths of data bus and a controller for the devices. The controller reduces unnecessary power consumption due to penetration power, etc. as much as possible. However, this conventional technique suffers from a low operational speed of a semiconductor memory device since the number of operations of a non-volatile memory device is large.
As NVDRAMs, the combination type NVDRAM including a DRAM as a volatile data storage means and an EEPROM as a non-volatile data storage means, and the ferroelectric type NVDRAM including a memory cell having volatile and non-volatile storage capabilities are described above. In either type NVDRAM, if only the non-volatile (recall/store) operations are performed continuously, a 10-year operation cannot be guaranteed.
Further, in the conventional NVDRAMs, there are problems in that the recall/store operations are slow; data may be not protected; and the polarization is fatigued.
In the NVDRAMs, data stored in the non-volatile state is transformed to data in the volatile state in the recall operation while data stored in the volatile state is transformed to data in the non-volatile state in the store operation. In the non-volatile (recall/store) operations, the polarization reversal of the ferroelectric material occurs, leading to the longer operation time as compared with during the volatile (DRAM) operation. Therefore, when the NVDRAM is operated in the non-volatile (recall/store) operations, there is a problem in that the read and write speeds are slow. There is also a problem in that a system noise or the like upon power-on or power-off may cause an erroneous rewrite operation which unexpectedly changes non-volatile data. The ferroelectric materials are fatigued after data rewrite operations utilizing the polarization reversal are repeated about 1010 to 1012 times. If the device is operated at a cycle of 100 nsec, a 10-year successive access operations to the memory cell cannot be guaranteed.
Further, when the NVDRAM has a large memory capacity, the refresh operation of the conventional volatile DRAMs increases power consumption.
The above-described problems are common to the case where the NVDRAM is implemented as a memory cell device solely; the case where the NVDRAM is implemented as a memory module or register for a single chip microcomputer (MPU or CPU); the case where the NVDRAM is incorporated into a system in the form of a card, and the like.
In the store operation of the combination type NVDRAM, a Fowler-Nordheim tunnel current flows through the EEPROM, requiring an additional time as compared with the volatile (DRAM) operation. Therefore, the write speed is slow. Moreover, the number of rewrite operations is limited to about one hundred thousand times for the EEPROM. The EEPROM will soon reach the end of its lifetime if frequent rewrite operation are conducted. A 10-year successive access operation thus cannot be guaranteed.
SUMMARY OF THE INVENTION
According to one aspect of the present invention, a non-volatile semiconductor memory device includes a memory cell array including memory cells capable of storing data in a volatile state and in a non-volatile state. The memory cells are arranged in a matrix. The non-volatile semiconductor memory device further includes a row address determination circuit for determining whether a row address in a current cycle is the same as a row address in a previous cycle when an activating signal for activating a read cycle for reading data stored in the memory cell or a write cycle for writing data into the memory cell is received. When the row address in the current cycle is different from the row address in the previous cycle, a store operation is performed with respect to the memory cell selected in the previous cycle, where volatile data is stored in the memory cell as non-volatile data; thereafter, a recall operation is performed where non-volatile data stored in the memory cell selected in the current cycle is transformed into volatile data; and the resulting volatile data is stored in the memory cell and a sense amplifier connected to a corresponding column address, or at least one of the sense amplifier and a latch circuit. When the row address in the current cycle is the same as the row address in the previous cycle, a read or write operation is performed with respect to volatile data stored in the memory cell, or the sense amplifier or latch circuit connected to a corresponding column address, but non-volatile data is not read from or written to the memory cell.
In one embodiment of this invention, when the read operation is performed immediately after power-on, a reset signal changing upon power-on or a reset signal inside the device is used in the recall operation.
In one embodiment of this invention, upon power-off or reset, a reset signal changing upon power-off or reset, or a reset signal inside the device is used in the store operation.
In one embodiment of this invention, the column address is assigned to lower address pin terminals, and the row address is assigned to higher address pin terminals; and an external address is input.
In one embodiment of this invention, each memory cell includes a combination of a DRAM as a volatile data storage means and an EEPROM as a non-volatile data storage means.
In one embodiment of this invention, each memory cell is made of a ferroelecric material; and data is stored in a non-volatile state in each memory cell using polarization reversal and data is stored in a volatile state in each memory cell without using the polarization reversal.
In one embodiment of this invention, the device is controlled in such a way that a driving signal for driving a word line connected to the memory cell is in the enable state during a period of time from the start of the recall operation to the end of the store operation.
In one embodiment of this invention, the device is controlled in such a way that a driving signal for driving a word line connected to each memory cell is in the enable state only during the recall operation and the store operation.
In one embodiment of this invention, the store operation is initiated after the recall operation is completed.
In one embodiment of this invention, the read or write cycle is performed in synchronization with an external synchronization signal.
In one embodiment of this invention, the non-volatile semiconductor memory device further includes a circuit for generating a WAIT signal, wherein the WAIT signal is inactive during the recall operation and the store operation.
In one embodiment of this invention, the non-volatile semiconductor memory device further includes a latch circuit for latching a row address at least during the read cycle and the write cycle.
According to another aspect of the present invention, a system LSI includes the non-volatile semiconductor memory device.
Hereinafter, functions of the present invention will be descried.
As NVDRAMS, the combination type NVDRAM including a DRAM as a volatile data storage means and an EEPROM as a non-volatile data storage means, and the ferroelectric type NVDRAM including a memory cell having volatile and non-volatile storage capabilities are described above. In either type NVDRAM, a memory cell has volatile and non-volatile data storage capabilities.
According to the present invention, in the NVDRAMs, when an activation signal is input f or activating a read cycle of reading data stored in a memory cell or a write cycle of writing data into the memory cell, the row address determination circuit determines whether a row address (word line address) input in a current read/write cycle is the same as a row address input in the previous read/write cycle.
When the row addresses differ from each other between the current and previous cycles, the store operation Is performed for a memory cell which has been selected by a row address in the previous cycle, where data stored in the volatile state is transformed into data stored in the non-volatile state. Thereafter, the recall operation is performed for a memory cell which is selected by a row address in the current cycle, where data stored in the memory cell in the non-volatile state is transformed into data in the volatile state for the purpose of reading. The obtained data stored in the memory cell in the volatile state is also stored in a sense amplifier connected to a column address corresponding to the memory cell. In either case of the ferroelectric type NVDRAM or the combination type NVDRAM, data may be stored only in the sense amplifier. Alternatively, data may be transferred and stored in the latch circuit rather than the sense amplifier. Alternatively, data may be stored in both the sense amplifier and the latch circuit. Data is stored in the volatile state in this way, thereby making it possible to reduce consumption of the polarization reversal. In the case of the combination type NVDRAM, the number of rewrite operations can be reduced. Further, the sense amplifier or the latch circuit includes a circuit which compensates a leakage current. Therefore, when data is stored in the volatile state in the sense amplifier or the latch circuit, the refresh operation which is performed in the conventional volatile DRAM is not required, thereby reducing power consumption.
When a row address selected in the current cycle is the same as a row address selected in the previous cycle, the read/write operations are performed with respect to data stored in the volatile state in a memory cell or a sense amplifier or latch circuit connected to a column address corresponding to the memory cell. In this case, the data stored in the non-volatile state is not read from or written to the memory cell. When the write operation is performed with respect to external data, if this operation is performed on the data stored in the volatile state in the sense amplifier or the latch circuit, the refresh operation which is performed in the conventional volatile DRAM is not required, thereby reducing power consumption. Further, when a column address selected in the current cycle is the same as a column address selected in the previous cycle, an operation requiring the polarization reversal (non-volatile operation) is not performed, so that the read speed is increased. Moreover, an SRAM having a small capacity can be omitted, thereby reducing cost. In the case of the combination type NVDRAM, the rewrite operation which causes the tunnel current is not performed, thereby increasing the rewrite speed.
When a device or system is turned on (powered on), a previous address is not specified. In this case, the recall operation can be controlled utilizing a reset signal which changes upon power-on (e.g., a system reset signal, or a reset signal within the device). Upon power-off, or when the system is reset, the store operation can be performed utilizing a reset signal which changes upon power-off or system reset (e.g., a system reset signal, or a reset signal within the device). At that point, by utilizing the system reset signal which inactivates the whole system, an erroneous write operation due to noise can be prevented.
For example, as will be described in the Examples, when the CE-bar (or the RAS-bar in the case of the address multiplex type) is activated (active state “LOW”) after power-on, the recall operation is performed with respect to a memory cell provided along a word line selected by a current row address. The resulting data obtained is then stored in the memory cell and a sense amplifier connected to the corresponding column address. Alternatively, the data may be stored in a latch circuit, or both in the sense amplifier and in the latch circuit. When a row address in a current cycle is the same as a row address in the previous cycle (a CEin-bar signal keeps the active (“LOW”) state), the read/write operation is performed with respect to data stored in the sense amplifier or the latch circuit connected to the corresponding column address. When a row address in a current cycle is different from a row address in the previous cycle (a CEin-bar signal keeps the inactive (“HIGH”) state), the store operation is performed with respect to data stored in the volatile state in the sense amplifier or the latch circuit is stored as data in the non-volatile state.
Whereas in the conventional NVDRAMs the recall operation is performed with respect to a memory cell using an NE (Non-Volatile Enable)-bar signal which enables a non-volatile operation, the recall operation is performed with respect to a memory cell selected by a row address when the CE-bar signal (or the RAS-bar signal in the case of the address multiplex type) is active (“LOW”) so that data is stored in the volatile state in a sense amplifier or the like which does not require the refresh operation and the other memory cells are used in the non-volatile manner which does not require the refresh operation. Thus, the refresh operation is not required for data stored in the volatile state, thereby reducing unnecessary power consumption. The CE-bar signal (or the RAS-bar signal) also serves as the NE-bar signal, so that a user does not recognize the NE-bar signal. Therefore, the NVDRAM of the present invention can be used in a way similar to that of the conventional volatile DRAMs, analogous SRAMs, or SRAMS, thereby obtaining satisfactory usability.
Further, the column address may be assigned to lower address pin terminals (A0-) and the row address may be assigned to higher address pin terminals. The address pin terminals are selected in ascending order from the lower address. This allows satisfactory compatibility with a system CPU and a control circuit, thereby obtaining satisfactory usability.
Typically, the number of column addresses having the same row address is about 100 to about 1000. The present invention is particularly effective to the case where data stored at serial addresses (e.g., file memory) are used. For example, the number of available polarization reversals (1012 to 1013) is multiplied by the number of the column addresses (100 to 1000), for exampe, 1000. The result is 1015 to 1016 which corresponds to the number of rewrite operations. The 10-year continuous operation (with a cycle of 100 nsec) can be achieved. In this case, the present invention is particularly effective when the column address is assigned to lower address pin terminals (A0-) and the row address is assigned to higher address pin terminals.
The NVDRAM of the present invention may be controlled in such a way that the driving signal (row decoder control signal) for driving a word line connected to a memory cell is in. the enable state (“HIGH”) during the period of time from the start of the recall operation to the end of the store operation. In this case, the circuit for generating the control signal can be simplified. Alternatively, the NVDRAM of the present invention may be controlled in such a way that the driving signal (row decoder control signal) for driving a word line connected to a memory cell is in the enable state only during the recall operation and the store operation. In this case, the driving signal is in the form of pulses, thereby reducing power consumption.
When the store operation is controlled upon power-off using a reset signal which changes upon power-off or system reset (e.g., a system reset signal or a reset signal within a device), the store operation can be initiated after the recall operation is completely ended so that data is not destroyed during the recall operation. An example of the reset signal within a device is a POWERGOOD signal.
Further, as in a synchronous system or a burst system, the read/write operations may be performed in synchronization with a synchronization signal from an external control system or the like. In this case, the synchronization signal can be utilized for the delay circuit so that the delay circuit can be simplified, thereby reducing variation in delay time.
A circuit for generating a WAIT signal may be provided so that the CPU or the control circuit can be in the inactive state (“LOW”) during the recall/store operations to keep the read/write states during the recall/store operations.
A row address may be latched when the CE-bar signal for controlling the read/write cycles is in the active state (active=“LOW”). This allows further prevention of an erroneous write operation due to noise or the like. The latching of a row address in such a period of time leads to prevention of an erroneous operation due to noise even when an input row address fluctuates due to noise, because the row address has already been latched inside the device.
Thus, the invention described herein makes possible the advantages of providing an NVDRAM and a system LSI which have high-speed recall/store operations and reduced power consumption and in which a 10-year continuous operation can be guaranteed.
These and other advantages of the present invention will become apparent to those skilled in the art upon reading and understanding the following detailed description with reference to the accompanying figures.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram showing an example of an NVDRAM according to the present invention.
FIG. 2 is a block diagram showing a configuration of a row address comparison/timing generation circuit used in the present invention.
FIG. 3 is a circuit diagram showing an example of a latch circuits for latching a row address and a comparison circuit used in the present invention.
FIG. 4A is a timing chart for explaining the operation of a ½ cycle circuit used in the present invention.
FIG. 4B is a circuit diagram showing an example of the ½ cycle circuit.
FIG. 4C is a circuit diagram showing an example of a pulse generation circuit which outputs the φ1, the circuit used in the present invention.
FIG. 5 is a circuit diagram showi ng an example of a CEin-bar signal generation circuit used in the present invention.
FIG. 6 is a circuit diagram showing another example of a CEin-bar signal generation circuit used in the present invention.
FIG. 7A is a block diagram showing an example of a row address latch circuit used in the present invention.
FIG. 7B is a timing chart for explaining the operation of the row address latch circuit.
FIG. 8A is a timing chart showing an example of a control signal waveform (word line driving form) for controlling the operation of the row decoder circuit upon the recall and store operations.
FIG. 8B is a circuit diagram showing an example of a circuit configuration for realizing the control signal waveform.
FIG. 9A is a timing chart showing another example of a control signal waveform (word line driving form) for controlling the operation of the row decoder circuit upon the recall and store operations.
FIG. 9B is a circuit diagram showing an example of a circuit configuration for realizing the control signal waveform.
FIG. 10 is a block diagram showing another configuration of an NVDRAM according to the present invention.
FIG. 11 is a diagram showing the hysteresis characteristic of a ferroelectric material.
FIG. 12 is a block diagram showing a configuration of a conventional NDVRAM including a memory cell including a ferroelectric material.
FIG. 13 is a diagram for explaining an operation in which data “0” is stored in the memory cell including the ferroelectric material in the store mode.
FIG. 14 is a diagram for explaining an operation in which data “1” is stored in the memory cell including the ferroelectric material in the store mode.
FIG. 15 is a diagram for explaining an operation in which data is read out from the memory cell including the ferroelectric material in the recall mode.
FIG. 16 is a block diagram showing an example of a conventional NVDRAM.
FIG. 17 is a block diagram showing another example of a conventional NVDRAM.
FIG. 18 is a block diagram showing an example of a conventional ferroelectric type NVDRAM.
FIG. 19 is a block diagram showing another example of a conventional NVDRAM.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the following embodiments, although NVDRAMs including a memory cell made of a ferroelectric material will be described, the NVDRAMs may be incorporated into a so-called system LSI or microprocessor unit (MPU) as a memory. The NVDRAMs may be used as a single memory device or a memory device for a one-chip microcomputer. Further, the NVDRAMs may be incorporated into a multi-chip module or the like which includes a plurality of devices, such as IC cards or PC cards which are standardized by the Personal Computer Memory Card International Association (PCMCIA) or the Japan Electronic Industry Development Association (JEIDA).
In the NVDRAMs of the present invention, a memory cell array may include a number of two-transistor/cell type memory cells shown in FIG. 12 which are immune to possible variations in the fabrication process. Alternatively, the memory cell array may include a number of one-transistor/cell type memory cells shown in FIG. 18 which are suitable for a higher degree of integration. The NVDRAM shown in FIG. 18 includes a dummy cell 41, an OR gate 42, a recall counter 43 for counting the number of recall operations, a switch 44 which performs ON/OFF switching for a recall operation when an extraordinary number of recall operations exceeding a predetermined number are successively input, a store counter 45 for counting the number of store operations, and a switch 46 which performs ON/OFF switching for a store operation when an extraordinary number of store operations exceeding a predetermined number are successively input.
Alternatively, as shown in FIG. 19, the NVDRAM may not include the NE-bar signal as a control signal. The NVDRAM shown in FIG. 19 includes a circuit 51 for generating φ1-bar, φ2-bar, and φ3-bar signals, logic gates 51a through 51h, a recall request latch circuit for receiving a recall completion signal from a self-recall mode timing control circuit 38a and a store completion signal from a self-store mode timing control circuit 39a, an address selection circuit 56 for selecting an external address or internal address, a driver/decoder circuit 57 for selecting and driving a WL line and a PT line, a memory cell array 58, and an access control circuit 59 for controlling the driver/decoder circuit 57 and the memory cell array. In FIG. 19, a CEi-bar indicates an i-th chip enable signal; an RFSH-bar indicates a refresh enable signal; an RF/RCi-bar indicates an i-th RE/RC-bar signal which indicates that a volatile operation is being performed when RF/RCi-bar=“THIGH”; and φ1-bar, φ2-bar, and φ3-bar indicate signals for controlling the timing control circuits 37, 38a, and 39a, respectively.
The self-recall mode timing control circuit 38a and the self-store mode timing control circuit 39a each include an internal address counter circuit and perform recall/store operations with respect to addresses in accordance with the internal addresses but not with external addresses.
Further, the present invention can also be implemented in the NVDRAMs including a DRAM as a volatile data storage means and an EEPROM as a non-volatile data storage means.
FIG. 1 is a block diagram showing a configuration of an NVDRAM according to an embodiment of the present invention. The NVDRAM includes a row address comparison/timing generation circuit 1, a store mode/recall mode timing generation circuit 2, a row address latch circuit 3, a volatile mode/address transition detection (ATD) timing generation circuit 4, a memory cell array 5, a row decoder 6, a sense amplifier 7, a column decoder 8, and an I/O latch and I/O output buffer circuit 9. Although the NVDRAM herein includes the volatile mode/address transition detection (ATD) timing generation circuit 4 which performs a read operation in synchronization with transition of addresses, the NVDRAM may include a so-called burst mode read circuit which can perform a read operation in synchronization with an external clock signal instead of an address signal.
In the NVDRAM, when the chip selection signal CE-bar transits from an inactive (non-selection) state to an active (selection) state, a read (recall) operation in which data stored in the non-volatile state is transformed to data stored in the volatile state is performed with respect to a memory cell selected by the corresponding row address. When the row address in the previous read/write cycle in the volatile mode is the same as the row address (word line address) in the current read/write cycle, the row address comparison/timing generation circuit 1 holds the CEin-bar signal in the “LOW” state (chip selection state). Thereafter, volatile data is read from or written to a sense amplifier connected to the corresponding column address. Non-volatile data is not read from a memory cell, and is not written to a memory cell. Therefore, the number of polarization reversals can be reduced. Moreover, when the volatile data is stored in the sense amplifier, the refresh operation is not required. When the volatile data is stored in the memory cell, the refresh operation is required, as the case of the conventional volatile DRAM.
When the row address in the previous read/write cycle in the volatile mode is different from the row address (word line address) in the current read/write cycle, the row address comparison/timing generation circuit 1 changes the CEin-bar signal from “LOW” to “HIGH” (from the chip selection state to the chip non-selection state). When the chip selection state changesto the chip non-selection state in this way, the store operation is performed with respect to a memory cell which has been selected by a row address in the previous cycle so that volatile data is transformed to non-volatile data. Thereafter, the CEin-bar changes from the non-selection state to the selection state, so that the recall operation is performed with respect to a memory cell selected by a row address in the current cycle. These recall and store operations can be performed in a way similar to that of the conventional NVDRAMs. It is here assumed that which address pin or layout lead line a row address (word line address) corresponds to is typically predetermined in accordance with a data sheet or circuit diagram. The volatile data obtained by the recall operation is stored in a sense amplifier connected to the corresponding column address. In this case, the refresh operation is not required. When volatile data is stored in a memory cell, the refresh operation is required.
The OE-barsignal is an output enable signal. The WE-bar signal is a write enable signal.
FIG. 2 is a block diagram showing a configuration of the row address comparison/timing generation circuit 1. The row address comparison/timing generation circuit 1 includes a latch circuit 11a for latching a first row address, a latch circuit 11b for latching a second row address, a comparison circuit 12 for comparing two row addresses, a CEin-bar generation circuit 13, a ½ cycle circuit 14 and pulse generation circuits 15a and 15b. The pulse generation circuits 15a and 15b generate latch control signals φ1 and φ2 at the odd-numbered falling and even-numbered falling of the control signal CE-bar, respectively. A row address externally input is latched as a first row address by the latch circuit llb in accordance with the latch control signal φ2. A row address externally input is latched as a second row address by the latch circuit 11b in accordance with the latch control signal φ2. The comparison circuit 12 compares the first address and the second address bit by bit. When the first and second addresses are the same, the output φ3 is “HIGH”. When the first and second addresses differ from each other, the output φ3 is “LOW”.
FIG. 3 is a circuit diagram showing an example of the latch circuits 11a and 11b for latching the row address and the comparison circuit 12. The comparison circuit 12 compares the first address and the second address bit by bit. When the first and second addresses are the same, the output φ3 is “HIGH”. When the first and second addresses differ from each other, the output φ3 is “LOW”.
FIG. 4A is a timing chart for explaining the operation of the ½ cycle circuit 14. FIG. 4B is a circuit diagram showing an example of the ½ cycle circuit 14. FIG. 4C is a circuit diagram showing an example of the pulse generation circuit 15a which outputs the φ1. The pulse generation circuit 15b which outputs the φ2 has a structure similar to that of the pulse generation circuit 15a, and receives an inverse signal of φA. The ½ cycle circuit 14 includes a D-type flip flop, which the output φA is preset to “HIGH” when a reset-bar signal is in the “LOW” state. A reset-bar terminal of the ½ cycle circuit 14 is connected to Vcc. The output φA from the ½ cycle circuit 14 is changed from “HIGH” to “LOW” or from “LOW” to “HIGH”, every time an input CE-bar signal changes from “HIGH” to “LOW”. The pulse generation circuit 15a outputs the φ1 having one pulse of “LOW” when the input φA changes from “LOW” to “HIGH”.
FIG. 5 is a circuit diagram showing an example of the CEin-bar signal generation circuit 13. The CEin-bar signal generation circuit 13 includes a first Delay circuit 16a, a second Delay circuit 16b, a third Delay circuit 16c, a fourth Delay circuit 16d, and D-type flip flops 17a and 17b.
When the reset bar signal (any one of the system reset signal and the reset signal inside the device) is “LOW”, the reset bar signal indicates that a system including the NVDRAM is in the reset state. In this case, the CEin-bar signal is “HIGH”, the internal signal φ4 is “HIGH”, and the internal signal φ5 is “LOW”.
A previous row address does not exist immediately after the system reset (including the case of Power On Reset immediately after power-on). Therefore, matching of the row addresses is forced only immediately after reset by the D-type flip flop 17a. Specifically, when the CE-bar signal is changed from “HIGH” to “LOW”immediately after reset (including the case immediately after power-on), the internal signal φ4 is changed from “HIGH” to “LOW” and the internal signal φ5 holds “LOW” after waiting times determined by the first and second Delay circuit 16a and 16b. In this case, the CEin-bar signal is changed from “HIGH” to “LOW”, and the recall operation is initiated with respect to a memory cell. selected by a row address. The first Delay circuit 16a is used to wait for a time more than or equal to a time which is required to compare row addresses. The second Delay circuit 16b is used to wait for a time more than or equal to a time which is required to establish the internal signal φ5.
The internal signal φ6 is initially “LOW”, because the reset bar signal is “LOW”. The WAIT signal which is inverse to the internal signal φ6 is “HIGH”. Thereafter, when the CE-bar signal is changed from “HIGH” to “LOW”, the signal φ6, which is inverse to the signal φ5 which is “LOW”, is “HIGH”. Therefore, the WAIT signal is changed from “HIGH” to “LOW” after a waiting time which isdeterminedbythefourthDelaycircuit 16d. Thefourth Delay circuit 16d causes the wait signal to be “HIGH” in the recall operation, and thus can instruct the CPU or the control circuit to hold the read/write states (not shown).
Thereafter, when a row address in a current cycle is the same as a row address in a previous cycle, the signal φ3 from the comparison circuit 12 is “HIGH”. Therefore, the internal signal φ5 and the CEin-bar signal hold “LOW” even when the CE-bar signal is changed.
When a row address in a current cycle is not the same as a row address in a previous cycle, the signal φ3 from the comparison circuit 12 is “LOW”. Therefore, the internal signal φ5 and the CEin-bar signal are changed from “LOW” to “HIGH” when the CE-bar signal is changed to “LOW”. In this case, the store operation is initiated with respect to a memory cell which has been selected by a row address in the previous cycle. Thereafter, the internal signal φ5 is changed from “HIGH” to “LOW” after a waiting time determined by the third Delay circuit 16c. Therefore, the waiting time of the third Delay circuit 16c is set to a time more than or equal to a time which is required for the store operation. The CEin-bar signal is also changed from “HIGH” to “LOW”, initiating the recall operation with respect to a memory cell selected by a current row address.
The WAIT signal is “HIGH” during the store and recall operations and outputs a WAIT request to the CPU and the control circuit.
When the fourth Delay circuit is omitted, the overall circuits can be simplified. In the synchronous type and the burst mode type, a synchronization clock signal is input from an external control system. A Delay circuit may count the number of the synchronization clock signals. In this case, the Delay circuit can be simplified and the synchronization can be reliably obtained, thereby reducing variation in delay time.
FIG. 1 shows a structure in which the resetbar signal is used. In this case, the reset-bar signal (e.g., RP (Reset Power Down) signal) is “LOW” upon power-on or power-off so that the chip is in the reset state. This leads to protection of data stored in the memory cell against system noise upon power-on or power-off. Upon power-on the reset-bar signal is “LOW”. The reset-bar signal is “HIGH” after a power supply reaches a predetermined voltage so that the chip is released from the reset state. Thereafter, when a row address is input, a previous row address does not exist, so that the recall operation is unconditionally performed. Upon power-off or system reset, the reset-bar signal may be set to “LOW” so that current consumption can be reduced. When the reset-bar signal is changed from “HIGH” to “LOW”, the chip is reset after the store operation is performed inside the device.
In this case, the store operation may be initiated after the recall operation is completed so that data is not destroyed during a period of time when the recall operation is performed inside the device. This can be realized by changing part of the circuit configuration shown in FIG. 5 to what is shown in FIG. 6. In the configuration shown in FIG. 6, the signal which is “HIGH” during the recall operation is generated by a latch circuit which, for example, is “HIGH” when the CEin-bar signal is changed from “HIGH” to “LOW”, and which is changed from “HIGH” to “LOW” when the recall operation is completed.
The reset-bar signal may be input through an external pin such as an existing RP-bar terminal. Alternatively, the reset-bar signal may be input by a command. Alternatively, the reset-bar signal may be input by applying a high voltage to a particular pin.
FIG. 7A is a block diagram showing an example of the row address latch circuit shown in FIG. 1. FIG. 7B is a timing chart for explaining the operation of the row address latch circuit. The row address latch circuit includes a row address latch circuit 18a, a row address latchcircuit 18b, a pulse generation circuit 19a and a pulse generation circuit 19b. When the control signal CE-bar is changed from “HIGH” to “LOW”, a pulse signal φR is generated and the row address latch circuit 18a is latched to a row address A. The control signal CEin-bar is changed from “HIGH” to “LOW”, a latch transfer pulse is generated in a pulse signal φx and the row address latch circuit 18b is latched to the row address A. The pulse signal φx is used for the recall operation with respect to a new row address B.
FIG. 8A is a timing chart showing an example of a control signal waveform (word line driving form) for controlling the operation of the row decoder circuit upon the recall and store operations. FIG. 8B is a circuit diagram showing an example of a circuit configuration for realizing the control signal waveform. Here, when the control signal CEin-bar is “LOW”, a row decoder control signal φy is “HIGH”. Upon power-on, the row decoder control signal φy is reset to “LOW” by a signal (“HIGH”) from a reset circuit and a signal (“HIGH” pulse) indicating store operation completion. According to this configuration, the circuit for generating the row decoder control signal can be simplified.
FIG. 9A is a timing chart showing another example of a control signal waveform (word line driving form) for controlling the operation of the row decoder circuit upon the recall and store operations. FIG. 9B is a circuit diagram showing an example of a circuit configuration for realizing the control signal waveform. Here, when the control signal CEin-bar is changed from “HIGH” to “LOW”, the row decoder control signal φy is changed from “LOW” to “HIGH”. When the recall operation is completed (a recall completion signal goes to “HIGH” and then reset), the row decoder control signal φy is changed from “HIGH” to “LOW”. When the control signal CEin-bar is changed from “LOW” to “HIGH”, the row decoder control signal φy is changed from “LOW” to “HIGH”. When the store operation is completed (a store completion signal goes to “HIGH” and then reset), the row decoder control signal φy is changed from “HIGH” to “LOW”. With this configuration, the row decoder is controlled by the pul se signal, thereby reducing power co nsumption.
FIG. 10 is a block diagram showing another configuration of the NVDRAM according to another embodiment of the present invention. In the configuration shown in FIG. 1, volatile data obtained by the recall operation is stored in the sense amplifier. In the configuration shown in FIG. 10, the volatile data is transferred and stored in a latch circuit 21 rather than the sense amplifier. The data read by the recall operation is transferred and stored to the latch circuit 21 via an opened transfer circuit (gate) 20. Upon the store operation, the transfer gate 20 is opened, so that data is transferred to the sense amplifier 7 and a memory cell from the latch circuit 21.
When the NVDRAM of the present invention is applied to another NVDRAM and a system, a CE-bar signal, an OE-bar signal, a WE-bar signal, an address signal, a data input signal, and a data output signal, which are required for control of the NVDRAM, are provided by a system control circuit (logic circuit), a system solution chip, a system CPU, or a system MPU. At least a chip selection signal, a read/write control signal, and a data signal are provided. In this case, the row address comparison/timing generation circuit 1 shown in FIG. 1 can be incorporated into a system control circuit (logic circuit), a system solution chip, a system CPU, or a system MPU.
In the NVDRAM, memory cells are accessed word line (row address) by word line. Hundreds to thousands of memory cells are connected to one word line (row address). For this reason, even when 8-bit data is read out, the recall operation (read operation of data stored in the non-volatile state) is performed word line by word line. Similarly, in the case of the write operation, the recall operation is performed with respect to data except for the write bits word line by word line.
Therefore, in the NVDRAM or a system using the NVDRAM, the row addresses in the memory cell array are assigned to higher bits of an address bus or address pin. This has more satisfactory usability than when the row addresses and column addresses for the memory cell array are randomly arranged in the address bus. In the latter case, an external system (e.g., memory card) is used so as to assign the addresses again (address scramble) in such a way that the row address is assigned to higher address pins and the column address is assigned to lower address pins. This leads to satisfactory matching between the addresses of the external system and the address inside the device, thereby obtaining more satisfactory usability.
When a number of the NVDRAMs or systems are used, each NVDRAM is selected by the CE-bar signal (CE1, . . . , CEn) decoded by an address. In only the selected NVDRAM, when the CE-bar signal is active (“LOW”), non-volatile data stored in a memory cell selected by a row address is transformed into volatile data in the recall operation. The resulting data is temporarily stored in a sense amplifier. Alternatively, the data is transferred from the sense amplifier to a latch circuit in which the data is temporarily stored. Thereafter, when a row address determination circuit (e.g., a row address comparison/timing generation circuit) for the NVDRAM selected upon the read or write operation determines that a row address in a current cycle is different from a row address in the previous cycle, the NVDRAM performs the recall operation. In this case, the WAIT signal output from the WAIT signal generation circuit is “HIGH”, requesting that a wait state is inserted into (i.e., the WAIT signal is applied to) a peripheral control circuit, an MPU, or the like. When row address determination circuit determines that a row address in a current cycle is the same as a row address in the previous cycle, the NVDRAM does not perform the recall operation. In this case, the WAIT signal output from the WAIT signal generation circuit is “LOW”, informing the peripheral control circuit, the MPU, or the like that the read operation can be performed immediately.
In the above-described embodiments, the non-synchronous NVDRAMs are described. The NVDRAM of the present invention may have a synchronous circuit configuration which receives a synchronous signal, such as a synchronous DRAM or a Rambus DRAM. In such a case, the synchronous signal can be utilized in the delay circuit, so that the configuration of the delay circuit can be simple, thereby reducing variation in delay time.
Further, the multiplex type NVDRAM can be provided in which the row address and the column address are divided in accordance with time and the divided addresses are taken in through the RAS-bar and the CAS-bar.
In addition to the above-described NVDRAM, various embodiments of the present invention can be provided, not departing from the viewpoints and methods of the present invention. For example, a non-volatile memory device using a tunnel magnetoresistance (TMR) film or a giant magnetoresistance (GMR) film can be used instead of the EEPROM. When the NVDRAM includes a plurality of memory cell arrays, the present invention can be applied to each memory cell array, thereby obtaining the similar effects. The above description and the drawings used herein do not limit the present invention in any manner.
As NVDRAMs, the combination type NVDRAM including a DRAM as a volatile data storage means and an EEPROM as a non-volatile data storage means, and the ferroelectric type NVDRAM including a memory cell having volatile and non-volatile storage capabilities are described above. In these NVDRAMs having a memory cell having volatile and non-volatile data storage capabilities, when an activation signal is input for activating a read cycle of reading data stored in a memory cell or a write cycle of writing data into the memory cell, the row address determination circuit can determine whether a column address (word line address) input in a current read/write cycle is the same as a column address input in the previous read/write cycle.
When the row addresses differs from each other between the current and previous cycles, the store operation is performed for a memory cell which has been selected by a row address in the previous cycle, where data stored in the volatile state is transformed into data stored in the non-volatile state. Thereafter, the recall operation is performed for a memory cell which is selected by a row address in the current cycle, where data stored in the memory cell in the non-volatile state is transformed into data in the volatile state for the purpose of reading. The obtained data stored in the memory cell in the volatile state is also stored in a sense amplifier connected to a column address corresponding to the memory cell, or the latch circuit rather than the sense amplifier, whereby the number of polarization reversal can be reduced. Further, when data is stored in the volatile state in the sense amplifier or the latch circuit, the refresh operation which is performed in the conventional volatile DRAM is not required, thereby reducing power consumption.
When a row address selected in the current cycle is the same as a row address selected in the previous cycle, the read/write operations are performed with respect to data stored in the volatile state in a memory cell or a sense amplifier or latch circuit connected to a column address corresponding to the memory cell. In this case, the data stored in the non-volatile state is not read from or written to the memory cell. This leads to a reduction in power consumption, and the operations using polarization reversal can be eliminated, thereby obtaining a high-speed read operation. When the write operation is performed with respect to external data, the data is written into the sense amplifier or the latch circuit is used, so that the refresh operation which is performed in the conventional volatile DRAM is not required, thereby reducing power consumption. Moreover, an SRAM having a small capacity can be omitted, thereby reducing cost.
When a device or system is turned on (powered on), a previous address is not specified. In this case, the recall operation can be controlled utilizing a reset signal which changes upon power-on (e.g., a system reset signal, or a reset signal within the device). Upon power-on, power-off, or when the system is reset, the recall and store operations can be performed utilizing a system reset signal which inactivates the whole system, whereby an erroneous write operation due to noise can be prevented.
For example, when the CE-bar (or the RAS-bar in the case of the address multiplex type) is activated (“LOW”) after power-on, the recall operation is performed with respect to a memory cell provided along a word line selected by a current row address. The resulting data obtained is then stored in the memory cell and a sense amplifier or a latch circuit connected to the corresponding column address. When a row address in a current cycle is the same as a row address in the previous cycle (a CEin-bar keeps the active (“LOW”) state), the read/write operation is performed with respect to data stored in the sense amplifier or the latch circuit connected to the corresponding column address. When a row address in a current cycle is different from a row address in the previous cycle (a CEin-bar keeps the inactive (“HIGH”) state), the store operation is performed with respect to data stored in the volatile state in the sense amplifier or the latch circuit is stored as data in the non-volatile state.
Whereas in the conventional NVDRAMs the recall operation is performed with respect to a memory cell using an NE (Non-Volatile Enable)-bar signal which enables a non-volatile operation, the recall operation is performed with respect to a memory cell selected by a row address when the CE-bar signal (or the RAS-bar signal in the case of the address multiplex type) is active (“LOW”) so that data is stored in the volatile state in a sense amplifier or the like which does not require the refresh operation and the other memory cells are used in the non-volatile mannerwhich does not requiretherefreshoperation. Thus, the refresh operation is not required for data stored in the volatile state, thereby reducing unnecessary power consumption. The CE-bar signal (or the RAS-bar signal) also serves as the NE-bar signal, so that a user does not recognize the NE-bar signal. Therefore, the NVDRAM of the present invention can be used in a way similar to that of the conventional volatile DRAMs, analogous SRAMs, or SRAMs, thereby obtaining satisfactory usability.
Further, the column address may be assigned to lower address pin terminals (A0-) and the row address may be assigned to higher address pin terminals. The address pin terminals are selected in ascending order from the lower address. This allows satisfactory compatibility with a system CPU and a control circuit, thereby obtaining satisfactory usability.
Typically, the number of column addresses having the same row address is about 100 to about 1000. The present invention is particularly effective to the case where data stored at serial addresses (e.g., file memory) are used. For example, the number of available polarization reversals (1012 to 1013) is multiplied by the number of the column addresses (100 to 1000). The result is 1015 to 1016 which corresponds to the number of rewrite operations. The 10-year continuous operation (with a cycle of 100 nsec) can be achieved.
Further, in a synchronous system or a burst system in which the read/write operations may be performed in synchronization with a synchronization signal from an external control system or the like, a high-speed operation can be realized and the configuration of the delay circuit can be simplified.
Various other modifications will be apparent to and can be readily made by those skilled in the art without departing from the scope and spirit of this invention. Accordingly, it is not intended that the scope of the claims appended hereto be limited to the description as set forth herein, but rather that the claims be broadly construed.
Claims (13)
What is claimed is:
1. A semiconductor memory device including:
a memory cell array including memory cells capable of storing data in a volatile state and in a non-volatile state, the memory cells arranged in a matrix; and
a row address determination circuit for determining whether a row address in a current cycle is the same as a row address in a previous cycle when an activating signal is received,
wherein when it is determined that the row address in the current cycle is different from the row address in the previous cycle, a store operation is performed with respect to the memory cell selected in the previous cycle, where data previously stored in a volatile state in the memory cell selected in the previous cycle is transformed into a non-volatile state in the memory cell selected in the previous cycle; thereafter, a recall operation is performed where data previously stored in a non-volatile state in the memory cell selected in the current cycle is transformed into a volatile state; and said data in the volatile state is stored in the memory cell and a sense amplifier connected to a corresponding column address, or at least one of the sense amplifier and a latch circuit; and
when the row address in the current cycle is the same as the row address in the previous cycle, a read or write operation is performed with respect to data in a volatile state stored in the memory cell selected in the current cycle, or the sense amplifier or latch circuit connected to a corresponding column address, but data in the non-volatile state is not read from or written to the memory cell selected in the current cycle.
2. A non-volatile semiconductor memory device according to claim 1, wherein when the read operation is performed immediately after power-on, a reset signal changing upon power-on or a reset signal inside the device is used in the recall operation.
3. A non-volatile semiconductor memory device according to claim 1, wherein upon power-off or reset, a reset signal changing upon power-of f or reset, or a reset signal inside the device is used in the store operation.
4. A non-volatile semiconductor memory device according to claim 1, wherein a column address is assigned to lower address pin terminals, and the row address is assigned to higher address pin terminals; and an external address is input.
5. A non-volatile semiconductor memory device according to claim 1, wherein each memory cell includes a combination of a DRAM as a volatile data storage means and an EEPROM as a non-volatile data storage means.
6. A non-volatile semiconductor memory device according to claim 1, wherein each memory cell is made of a ferroelecric material; and data is stored in a non-volatile state in each memory cell using polarization reversal and data is stored in a volatile state in each memory cell without using the polarization reversal.
7. A non-volatile semiconductor memory device according to claim 1, wherein the device is controlled in such a way that a driving signal for driving a word line connected to the memory cell is in the enable state during a period of time from the start of the recall operation to the end of the store operation.
8. A non-volatile semiconductor memory device according to claim 1, wherein the device is controlled in such a way that a driving signal for driving a word line connected to each memory cell is in the enable state only during the recall operation and the store operation.
9. A non-volatile semiconductor memory device according to claim 1, wherein the store operation is initiated after the recall operation is completed.
10. A non-volatile semiconductor memory device according to claim 1, wherein the read or write cycle is performed in synchronization with an external synchronization signal.
11. A non-volatile semiconductor memory device accordi ng to claim 1 further Includes a circuit for generating a WAIT signal, wherein the WAIT signal is inactive during the recall operation and the store operation.
12. A non-volatile semiconductor memory device according to claim 1 further includes a latch circuit for latching one of the row address in the current cycle and the row address in the previous cycle at least during the read cycle and the write cycle.
13. A system LSI including a non-volatile semiconductor memory device according to claim 1.
US09/656,3142000-01-212000-09-05Non-volatile semiconductor memory device and system LSI including the same
ActiveUS6370058B1
(en)
Free format text:
CORRECTIVE TO CORRECT THE FILING DATE FROM 09-06-2000 TO 09-05-2000 PREVIOUSLY RECORDED AT REEL 011417, FRAME 0770 THE ASSIGNOR HEREBY CONFIRMS THE ASSIGNMENT OF THE ENTIRE INTEREST.;ASSIGNOR:FUKUMOTO, KATSUMI;REEL/FRAME:011728/0725
|
<?xml version="1.0" encoding="UTF-8"?>
<form xmlns="http://www.intellij.com/uidesigner/form/" version="1" bind-to-class="org.elixir_lang.credo.Configurable">
<grid id="27dc6" binding="panel" layout-manager="GridLayoutManager" row-count="1" column-count="1" same-size-horizontally="false" same-size-vertically="false" hgap="-1" vgap="-1">
<margin top="0" left="0" bottom="0" right="0"/>
<constraints>
<xy x="20" y="20" width="500" height="400"/>
</constraints>
<properties/>
<border type="none"/>
<children>
<component id="e65ee" class="javax.swing.JCheckBox" binding="includeExplanationCheckBox">
<constraints>
<grid row="0" column="0" row-span="1" col-span="1" vsize-policy="0" hsize-policy="3" anchor="8" fill="0" indent="0" use-parent-layout="false"/>
</constraints>
<properties>
<text value="Include Explanation"/>
<toolTipText value="Disabling the explanation tool tip eliminates an extra call to `mix credo PATH:LINE[:COLUMN]` for each issue, so it can improve the performance significantly. Uncheck if `mix credo` is spiking your CPU. If your CPU is still spiking when analyzing individual files, you can disable the Credo annotator by turning off the inspection: 1. Preferences > Editor > Inspections > Elixir > Credo 2. Uncheck the box."/>
<verticalAlignment value="1"/>
<verticalTextPosition value="0"/>
</properties>
</component>
</children>
</grid>
</form>
|
Turkish woman tells how brother and son fled to ISIL’s stronghold in Syria
Report: İpek İZCİ Photos: İbrahim YURTBAY Translator: Yasemin GÜLER
A 37-year-old housewife, born and raised in the central Anatolian province of Konya... She wants everyone to hear what she has to say, with the condition that we keep her name hidden from the public. First, her brother took his wife and children and left to join the Islamic State of Iraq and the Levant (ISIL) in its Syrian stronghold of al-Raqqah. Then, her 18 year old son followed in his uncle’s footsteps…As the first American drones taking off from Turkey started to pound ISIL targets in al-Raqqah on Aug. 5, the Turkish housewife is terrified that she will lose her youngest son to ISIL as well - as would any mother be.She was blamed as being an infidel by her son, for whom she recited the Quranic verse of Ayat Al-Kursi every time he left the house, in the hope of keeping him safe from harm. “Their return would be of no use. Their brains have completely frozen over. They don’t accept me or anyone else,” she says.My brother left on the October 11, 2014. We found out about it four days later. He just up and left with his entire family.35.Of course. He used to go to ISIL gatherings frequently before he left.In the district of Karatay. There’s a place they all meet up on Saturdays. Some 300 people gather and talk.It gets so crowded that the shoes form a small hill on the side of the road.He began attending meetings there two years ago. He married a like-minded girl and they ran away silently.These people that join ISIL call themselves Muslims; she wouldn’t even visit her own mother because she didn’t think she was a “proper” Muslim. She would only visit her mother-in-law. Because those that aren’t with ISIL are infidels to them. Her first husband died in Syria but she didn’t learn from that. Then she married my brother and went back there.They had left a hand-written letter: “We are leaving, to live the way Allah tells us to and to eventually become martyrs.”My sister-in-law has a son from her previous marriage and a 10-year-old son with my brother. She also had another baby there.No, none at all. They brainwashed him at those meetings.He would play a DVD every time he came over and make us watch. He would say, “You are infidels,” and he would tell us not to vote.For example, on one of them, someone would be reading the Quran and a woman would pop up every now and then and say, “Those who play ball become sinners.” They don’t approve of the motto, “I am Turkish, I am true, I am hardworking,” and they scold and talk down at Atatürk. They watch that all day. When my mother went to their house they would have it on all day.They used to call my mom, but since my son has gone there, no one has heard anything from them.He is 18. He left on May 18. Silently, just like his uncle… I was ill, at home. He came home from work at 10. “What happened darling, you’re early,” I said. He was very excited. He told me he had forgotten his work clothes. He took them and left. I looked at him at as he left because I used to recite the Quranic verse of Ayat Al-Kursi after him. Turns out that that was the last time I would see him.He was a placid young man. Very hygiene obsessed. If I touched my face while cooking, he would tell me to wash my hands. He was a baker. He used to bake local bread that we call “etli ekmek” (bread with meat). He wouldn’t harm anyone. The only thing I didn’t approve of was the smoking; he was a nicotine addict. His uncle had posted some things on the Internet after he had gone to Syria. There were photos of him sitting at his house, drinking tea. They live in looted homes over there. Their homes and clothes are so luxurious and expensive-looking, my son was influenced by them all. He watched ISIL on the Internet for two months straight.“Darling, don’t look at them,” I would say but he still would. He left, all the while staring at the butchers and barbarians… I tried my best, I took him to elders, to hodjas; I went all the way to the mufti’s office. The mufti would talk to my son but he wouldn’t listen. He became completely brainwashed once he began going to those gatherings. He had been in contact with his uncle in secret; I had no idea. The day he left home, that was the first place I looked for him but it was closed.I told them that he must still be here and to find him. “No, he must have already gone,” was the reply I got. “But, he will call me and ask me for my blessing.” Those who go to the border do that, which is why I thought he would call.He sent one of his friends a voice-message at the border. I told him to tell my son to wait for him and pretend that he wanted to join him in Syria. So we could buy some time and catch him before he crossed the border…“Look, you’re not joking, are you?” he asked his friend. “I can get someone to pick you up from [the southeastern Turkish province of] Gaziantep, I’m in al-Raqqah at the moment.” But he had told me that he was going to his father’s house in the evening.This is my second marriage. His father lives in Konya. Turns out that day he went to Gaziantep. They picked him up from there and took him to al-Raqqah.My son had given the name and phone number of the man in al-Raqqah to his friend over the phone. We shared this information with anti-terror police but they told us that this wasn’t a crime.They took interest and helped as much as they could but told us that they couldn’t do anything if he had left Turkish borders. “Even if we managed to catch him while he was still on Turkish soil, there wouldn’t be anything we could do. He is 18, he has the freedom to travel,” the police said.Just once… The elections were on June 7, and we spoke a week later. “My darling boy, I voted today.[Recep] Tayyip Erdoğan won again, the constitution will change,” I told him. “Have you become an infidel, mother? Why did you vote, why did you sin? This place is messed up because of him. You are not a Muslim,” he said to me. “A Muslim shouldn’t kill another Muslim, my son,” I said. “Mom, this place isn’t that sort of place; we kill the PKK [the outlawed Kurdistan Workers’ Party].”His voice kept cutting out. He told me he was working at a baker’s at the time. I told him not to come back unless he changed the way he thought.“Just because you said that now, I’m going to go to war,” was his reply. And he did. He is waiting guard at a front in a village somewhere. But recently he spoke to one of his friends over the Internet. He told him that he, “regret[s] coming here, [he] can’t smoke.” His friend asked him what the point of going was if he was going to return. He told him that he was afraid of going to jail. He had told me that too.Ahmet Mahmut Ünlü [a Turkish Sunni televangelist] says, “There are many that regret what they have done, but how will they return?” And now that he has told his friend that, I’m wondering whether my child will be tracked from the Internet. Because his friend asked him on Facebook what he was talking about and that he hadn’t committed a crime. He hasn’t got a response since.For the jihad.He earned well. He had everything and more.No, there wasn’t anyone. He went to work and came back home, that was it.He went alone.Yes, he’s 16. He is a baker too. I’m worried about him as well. His uncle took him to one of the gatherings while he was still here. “All of the men there looked like the Devil to me, mom. I felt suffocated,” he told me.Hopefully he won’t go. Those that go become suicide bombers and go to war. All these religious communions and gatherings have to be stopped. But the anti-terror police claim, “It’s not a crime. On what grounds can we shut them down?”What could he do? What could he have done?No, he drinks alcohol. That’s why we broke up in the first place.Sometimes I think to myself, “He went on his own, it was his own decision. There’s nothing I can do.” That’s how I console myself. But deep down I know that he had been tricked. My own brother, my flesh and blood, tricked him. He used to sit on this sofa, when he came home from work [She begins crying]. I’m supposed to let you run away with terrorists after raising you, am I?My brother has a 10-year-old son. He was glad to hear that my son had arrived in Syria. “Oh, my big brother has become a Muslim too,” he apparently said. A boy at that age… They pick whatever suits them from the Quran and dismiss the rest. Their return will be of no use… Their brains have frozen over. They don’t accept me or you. They’re extremely arrogant. “Son, those who have even a pinch of arrogance in them won’t go to Heaven,” I told him. But they have an answer for everything.
|
The described invention relates in general to an obstacle detection system for use with vehicles, and more specifically to an overhead obstacle detection and avoidance system that includes an ultrasonic sensor and processing unit for determining the height of an overhead obstacle and comparing that height to the maximum height of a vehicle.
Large trucks and tractor-trailer units are commonplace in modern society and are used on a daily basis for transporting a multitude of items from one place to another. The highway system in the United States as well as in most other countries includes numerous bridges and overpasses that may create difficulties for the operators of trucks and tractor-trailer units due to the fact that the height of some vehicles exceeds the clearance required to safely travel under these obstacles. If a truck or tractor-trailer unit exceeds the required clearance of a bridge or overpass, and the operator of the vehicle is not able to make that determination prior to encountering the obstacle, a collision will occur. These collisions often result in extensive damage to both the vehicle and the section of the roadway involved. The operator of the vehicle, the operators of other vehicles on the affected section of road, and even pedestrians crossing a bridge or overpass may also be injured or even killed.
Despite the problems described above, few systems for allowing vehicles such as trucks or tractor-trailer units to avoid collision with overhead obstacles are commercially available. Many previously developed systems are either prohibitively expensive to install, are prohibitively difficult to use, or are only somewhat effective for their intended purpose. Thus, there is an ongoing need for a reliable, relatively inexpensive and easy to install detection system for allowing large (i.e., tall) vehicles to avoid costly and dangerous collisions with overhead obstacles such as bridges and overpasses.
|
Steve's Website
SIMPLE
here is an article from my friend Doug Roe
for link to a great new servezine check out http://serve-others.com/issue2
How Do You Spell Effective?...S-I-M-P-L-EBy Doug Roe
A few years ago, I was standing on a street corner giving away cans of soda. Traffic backed up as people stopped to get their free soda. While most people were pleasantly surprised, one man stood out as particularly unhappy. He pulled to the corner, the windows of his big dark blue Buick rolled up, with an annoyed look on his face.
“Diet or regular?” I asked. He took a diet soda – though he was not appeased.
Several weeks later I was washing cars at a business near the same intersection when who should appear? The same annoyed man in the big Buick!
Annoyed Man!You are causing a traffic jam!My job that day was to be the “wheelman” – that is, to wash tires and whitewalls. I was on my hands and knees scrubbing the grim from the annoyed man’s tires, when I looked up to see him towering over me like a vulture. He had the same annoyed look on his face. I figured that I needed to say something before he did.
The first thing that popped into my head was this: “If Jesus were here, what would you ask of him?”
The question seemed to disarm him, and for a moment he forgot that my act of kindness had annoyed him. He offered money for the car wash, but I refused. I continued my offer to pray for him. He asked me to pray for his family, so I prayed a quick prayer and he was on his way.
A few weeks later, low and behold, I was shocked to see “Mr. Buick” and his family seated on the front row at our Easter Sunday service. At the end of the service, he and his wife came forward and asked me to pray with them. We prayed, and he accepted Jesus as his Savior.
In more than 20 years of serving others, I have learned a few things that have helped me enjoy the ride. * Servant Evangelism cannot be delegated!
People will follow what they see their leaders doing. Leader must be involved in serving others as a completely natural lifestyle – an overflow of their living, vital relationship with Jesus. If you are the lead pastor you don’t have to be the one developing the ideas and doing all of the legwork, but you do have to be the lead server.
* Start small with out-of-the-box ideas.
Often when we start a new project we think of starting something that is creative and dazzling. The power of serving is not in the size or “uniqueness” of the project. Projects that capture the inspired, creative attention of your people don’t need to be “sexy” or have a lot pop and sizzle. Simplicity if often the best approach to take. If you get nothing else in this article get this: The essence of the power of serving is in showing up in unexpected places. It is a simple act of compassion shown in a place where it is completely unexpected. Keep it simple, and the creative things will follow.
* Listen to your community.
Every community is different. And those unique needs that can and will be met by the unlimited resources of the kingdom of God flowing in and around us. The key? To find a need and then meet that need. The best ideas for reaching the people in your city are to listen to the people in your city – understanding their rhythms and needs.
While shopping at a local grocery store, I overheard two single mothers discussing what groceries they had put back in order to get the oil changed in their cars. The thought popped into my head, “I can do that! I can help them change oil in their car!”
So I went to a local mechanic to ask if he would be willing to help. “Now that is what the church should be about!” he said, blowing a puff of cigarette smoke in my face. We now change the oil in 300-400 cars a year, and we are beginning a new, broader ministry to single parents.
* Servant Evangelism is like growing asparagus.
When thinking about evangelism, we often think of wheat, but I have discovered that it is more like growing asparagus. When a gardener grows asparagus it takes three years before he can yield a crop.Asparagus Bunches
Pastors often ask, “If I do this serving thing in my church how many people will come to my church?” That is the wrong question to ask, and really reflects a “me” mentality. Servant evangelism is about the Church being the Church and serving others. If I try to grow my local church things just get mucked up. But if I serve other people, seek to expand the kingdom of God, and do the deeds that Jesus did I find that God brings the people. In the end it is not about the size of our congregation or budget, but about the number of people we have loved along the way.Gene, the annoyed man in the Buick, has been part of the church for more than 10 years now. He and his wife lead a small group. This once annoyed driver now helps me give soda away to other annoyed drivers.
“You know that day I stepped out of the car while you were washing the tires?” he asked one day.
“Yes,” I said, very clearly remembering it.
“I was going to chew you out for having caused such a traffic jam the day you gave out the sodas,” he said.
“What stopped you?” I asked.
“The fact that you wanted to pray for me,” he said. “I still have the can of diet soda you gave me sitting on the mantel of my fireplace. I kept it because it was the starting point that brought me to Christ.”
People may not flood into your church because you gave them a soda, but their lives will be changed because you were willing to share God’s love with them.
Doug Roe
Doug Roe is the Founding Pastor of the Dayton Vineyard, “A pretty good church.” Doug planted the Vineyard 15 years ago. Doug’s passion is to reach broken, hurting people from all walks of life, restoring dignity and empowering people to do ministry. His life mission is “to leave Dayton better than I found it.” In Doug’s approach to ministry, he strives to be both spiritual and smart at the same time. With the help of some good friends, the Dayton Vineyard has grown from a few people to regular attendance of more than 2,300 per week.
|
Universal to remove anti-piracy technology
August 9, 2007 6:01 PMMatt Marshall
The Universal Music Group, the world’s largest music label, plans to sell a significant portion of its catalog without copy protection software for at least the next few months, according to the New York Times.
It’s one more sign that the industry’s united stance behind digital-rights management anti-piracy technology may be falling apart.
Universal will offer the rights-free music through retail services like RealNetworks, Wal-Mart, Amazon.com, Google, and other sites, the Times said, but won’t be offered through Apple’s iTunes — a possible sign that Universal wants to compete with iTunes, or at least shift momentum away from Apple which has become the largest, most powerful music service.
Previously, EMI Group moved away from digital-rights, and signed a deal with iTunes. Under that agreement, though, songs are sold at a higher price — $1.29 instead of 99 cents.
|
Effect of ether- and water-soluble fractions of Carica papaya ethanol extract in experimentally induced hyperlipidemia in rats.
The papaya is the fruit of the plant Carica papaya L. (Caricaceae) used in India. Fruit and latex are both rich in an enzyme called papain. It is used as a folk remedy for contraception and abortion. The present study explored the anti-hyperlipidemic effect of the ether- and water-soluble fractions of C. papaya ethanol extract in olive oil-induced hyperlipidemic rats. The study also involved chromatographic studies of extract and fractions. Flash chromatography was done for the most active fraction. The extract and fractions were administered orally at doses of 200 and 400 mg/kg body weight in rats. Olive oil (5 mL/kg oral dose) was administered 30 min after treatment. Blood was collected and centrifuged at 3000 rpm for 15-20 min and subjected to biochemical analysis. The study dose-dependently inhibited the total cholesterol (TC), triglycerides (TG), low-density lipoproteins (LDL) level, and significantly increased high-density lipoprotein (HDL) level. Phytochemical screening revealed the presence of fats in the ether fraction, whereas the water fraction revealed the presence of tannins, alkaloids, glycosides. UV λ(max) was found to be 217 nm with a melting point of 41°C for the isolated component. The anti-hyperlipidemic effect was evaluated in olive oil-loaded rats. Acute treatment caused stimulatory effect on HDL level and inhibition in TC and TG elevation induced by olive oil. The extract and water fraction showed protective action by increasing the HDL cholesterol level.
|
Ok at your weight and with the Glide and surfboard, you will probably be fine with the 12 unless you are really trying to get out in real light stuff. I'll bet the new Rebel 12 has more grunt and a wider range than the Contra. Probably turns quicker too and no pulleys! The '08 bar is simpler than the '07. Nice lines, resist tangles. Better cleat and chicken loop line.
I'm 104kg and ride a 14m and a 10m with a 135cm 08 x-ride and a surfboard for really light winds. I am a little spoilt though because my girlfriend got ride of her switchblade 2's and got a 12m and a 9m (60kg or so never asked). I ride the 14 from 10-20 and the 10 from 20-35. there's an overlap or 5 knots 20-25 where I can ride either kite.
Can't say enough good things about the 08 rebels love'm. Never tried the 07's and still need to try out the back pig tail conection points yet and I've been using the shorter chicken line through and more depower strap. More kites the better.
Give those rear pigtails a try. I doubt you will go back to the forward position after trying them. Cuts the bar pressure in half and increases the turning speed. The only thing is you loose a little bottom end, but you can make it up by flying the kite activly.
I agree with the rear pigtails on my '08 14 is the best. I did find with the stock setup on the '08 bar that the kite flew on the front lines too much and could stall and fall back. Lengthed the back lines an inch and found that now the kite flews beautifully. Anyone else try that?
Give those rear pigtails a try. I doubt you will go back to the forward position after trying them. Cuts the bar pressure in half and increases the turning speed. The only thing is you loose a little bottom end, but you can make it up by flying the kite activly.
I also notice that my 7 Rebel 08 had a tendency to stall when the bar was fully sheeted in. So I also lengthened the rear lines and now is perfectly trimmed!
Hello dksurfer,
Do you know what's the wind range for the 7 sm Rebel 08 ? what is your wieght ? I have a north rhino 12 m from last year, do you thing it will pass together ? what's the low end of the rebel 7 m ? what's the high end ? thanks a lot. I am about to get a 7 m Rebel.
Who is online
Users browsing this forum: No registered users and 5 guests
You cannot post new topics in this forumYou cannot reply to topics in this forumYou cannot edit your posts in this forumYou cannot delete your posts in this forumYou cannot post attachments in this forum
|
Q:
Communication between different C# based services
Is there a way to communicate between two different services? I have a service that already runs. Is there a way to create a second service that can attach to the first service and send and receive dates to it?
I would also like to access the Windows service from a console application and attach to it. Is it possible?
A:
You can try to implement this by using:
IPC (Inter Process Communication via Named pipes)
Shared memory (Memory mapped files)
Socket (TCP/IP)
Example of using WCF: Many to One Local IPC using WCF and NetNamedPipeBindin.
Other example: A C# Framework for Interprocess Synchronization and Communication.
Everything depends on what version of .NET Framework you use. If you use .NET 3.0 and above then you can take a look into WCF. If not then you are on your own and you can google on keywords P/Invoke (CreateFileMapping, MapViewOfFile, CreatePipe...).
A:
To begin with I would play around with tcpclient and tcpserver
http://msdn.microsoft.com/en-us/library/system.net.sockets.tcpclient.aspx
http://msdn.microsoft.com/en-us/library/system.net.sockets.tcplistener.aspx
Even if the data you need to send is more complex than a date it can easily be serialized/deserialized.
For sending and receiving dates this seams the simplest option.
Also socks work if the services run on different machines whereas shared memory and namedpipes don't.
example code
// Create a thread running this code in your onstarted method of the service
using System.IO;
using System.Net;
using System.Net.Sockets;
var server = new TcpListener(IPAddress.Parse("127.0.0.1"), 8889);
server.Start();
while(true) {
var client = server.AcceptTcpClient();
using(var sr = new StreamReader(client.GetStream())) {
var date = DateTime.Parse(sr.ReadToEnd());
Console.WriteLine(date);
}
}
// In the console
using System.IO;
using System.Net;
using System.Net.Sockets;
var client = new TcpClient("localhost",8889);
using(var sw = new StreamWriter(client.GetStream())) {
sw.Write(System.DateTime.Now);
}
|
<!doctype html>
<html>
<head>
<title>Test'em</title>
<script src="/testem.js"></script>
<script src="//code.jquery.com/qunit/qunit-1.20.0.js"></script>
<script>
QUnit.config.autostart = false;
setTimeout(function() {
QUnit.start();
}, 250);
Testem.hookIntoTestFramework();
</script>
<script src="error.js"></script>
<link rel="stylesheet" href="//code.jquery.com/qunit/qunit-1.20.0.css"/>
</head>
<body>
<div id="qunit"></div>
<div id="qunit-fixture"></div>
</body>
</html>
|
/*
* Copyright (C) 2010-2015 Nektra S.A., Buenos Aires, Argentina.
* All rights reserved. Contact: http://www.nektra.com
*
*
* This file is part of Deviare
*
*
* Commercial License Usage
* ------------------------
* Licensees holding valid commercial Deviare licenses may use this file
* in accordance with the commercial license agreement provided with the
* Software or, alternatively, in accordance with the terms contained in
* a written agreement between you and Nektra. For licensing terms and
* conditions see http://www.nektra.com/licensing/. Use the contact form
* at http://www.nektra.com/contact/ for further information.
*
*
* GNU General Public License Usage
* --------------------------------
* Alternatively, this file may be used under the terms of the GNU General
* Public License version 3.0 as published by the Free Software Foundation
* and appearing in the file LICENSE.GPL included in the packaging of this
* file. Please visit http://www.gnu.org/copyleft/gpl.html and review the
* information to ensure the GNU General Public License version 3.0
* requirements will be met.
*
**/
#pragma once
//-----------------------------------------------------------
template <class T, typename _uuid>
class ATL_NO_VTABLE IDynamicConnectionPoint : public _ICPLocator<&__uuidof(_uuid)>
{
typedef CComEnum<IEnumConnections, &__uuidof(IEnumConnections), CONNECTDATA,
_Copy<CONNECTDATA>> CComEnumConnections;
public:
IDynamicConnectionPoint() : _ICPLocator<&__uuidof(_uuid)>()
{
return;
};
~IDynamicConnectionPoint()
{
SIZE_T i, nCount;
nCount = aList.GetCount();
for (i=0; i<nCount; i++)
{
if (aList[i].lpStream != NULL)
{
::CoReleaseMarshalData(aList[i].lpStream);
aList[i].lpStream->Release();
}
if (aList[i].lpObj != NULL)
aList[i].lpObj->Release();
}
aList.RemoveAllElements();
return;
};
STDMETHOD(_LocCPQueryInterface)(__in REFIID riid, __deref_out void **ppvObject)
{
ATLASSERT(ppvObject != NULL);
if (ppvObject == NULL)
return E_POINTER;
*ppvObject = NULL;
if (InlineIsEqualGUID(riid, __uuidof(IConnectionPoint)) || InlineIsEqualUnknown(riid))
{
*ppvObject = this;
AddRef();
return S_OK;
}
return E_NOINTERFACE;
};
STDMETHOD(GetConnectionInterface)(__out IID* piid2)
{
if (piid2 == NULL)
return E_POINTER;
*piid2 = __uuidof(_uuid);
return S_OK;
};
STDMETHOD(GetConnectionPointContainer)(__in IConnectionPointContainer** ppCPC)
{
T* pT = static_cast<T*>(this);
// No need to check ppCPC for NULL since QI will do that for us
return pT->QueryInterface(__uuidof(IConnectionPointContainer), (void**)ppCPC);
};
STDMETHOD(Advise)(__in IUnknown *pUnkSink, __out DWORD *pdwCookie)
{
T* pT = static_cast<T*>(this);
ITEM *lpItem;
TNktComPtr<IStream> cStream;
TNktComPtr<IDispatch> p;
HRESULT hRes;
if (pdwCookie != NULL)
*pdwCookie = 0;
if (pUnkSink == NULL || pdwCookie == NULL)
return E_POINTER;
hRes = pUnkSink->QueryInterface(IID_IDispatch, (void**)&p);
if (FAILED(hRes))
return (hRes == E_NOINTERFACE) ? CONNECT_E_CANNOTCONNECT : hRes;
//create marshaller
hRes = ::CreateStreamOnHGlobal(NULL, TRUE, &cStream);
if (SUCCEEDED(hRes))
hRes = ::CoMarshalInterface(cStream, IID_IDispatch, p, MSHCTX_INPROC, NULL, MSHLFLAGS_TABLESTRONG);
if (SUCCEEDED(hRes))
{
CNktAutoFastMutex cLock(&cMtx);
SIZE_T i, nCount;
//find a free slot and add to list
nCount = aList.GetCount();
for (i=0; i<nCount; i++)
{
if (aList[i].lpObj == NULL)
break;
}
if (i >= nCount)
{
lpItem = aList.ReserveBlock(1);
if (lpItem != NULL)
{
lpItem->lpObj = p.Detach();
lpItem->lpStream = cStream.Detach();
}
else
{
hRes = E_OUTOFMEMORY;
}
}
else
{
aList[i].lpObj = p.Detach();
aList[i].lpStream = cStream.Detach();
}
*pdwCookie = (DWORD)(i+1);
}
if (FAILED(hRes))
::CoReleaseMarshalData(cStream);
return hRes;
};
STDMETHOD(Unadvise)(__in DWORD dwCookie)
{
T* pT = static_cast<T*>(this);
TNktComPtr<IDispatch> p;
TNktComPtr<IStream> cStream;
LARGE_INTEGER li;
{
CNktAutoFastMutex cLock(&cMtx);
if (dwCookie < 1 || (SIZE_T)dwCookie > aList.GetCount())
{
NKT_ASSERT(FALSE);
return CONNECT_E_NOCONNECTION;
}
dwCookie--;
p.Attach(aList[dwCookie].lpObj);
cStream.Attach(aList[dwCookie].lpStream);
aList[dwCookie].lpStream = NULL;
aList[dwCookie].lpObj = NULL;
}
if (cStream != NULL)
{
li.QuadPart = 0;
cStream->Seek(li, SEEK_SET, NULL);
::CoReleaseMarshalData(cStream);
}
return S_OK;
};
STDMETHOD(EnumConnections)(__deref_out IEnumConnections** ppEnum)
{
T* pT = static_cast<T*>(this);
CComObject<CComEnumConnections>* pEnum;
TNktComPtr<IStream> cStream;
TNktComPtr<IUnknown> cUnk;
TNktComPtr<IDispatch> p;
LARGE_INTEGER li;
CONNECTDATA *lpCd;
SIZE_T i, k, nItems, nCount;
HRESULT hRes;
if (ppEnum == NULL)
return E_POINTER;
*ppEnum = NULL;
pEnum = NULL;
ATLTRY(pEnum = new CComObject<CComEnumConnections>)
if (pEnum == NULL)
return E_OUTOFMEMORY;
{
CNktAutoFastMutex cLock(&cMtx);
nCount = aList.GetCount();
lpCd = new CONNECTDATA[(nCount > 0) ? nCount : 1];
if (lpCd == NULL)
{
delete pEnum;
return E_OUTOFMEMORY;
}
//duplicate streams
hRes = S_OK;
for (i=nItems=0; i<nCount && SUCCEEDED(hRes); i++)
{
if (aList[i].lpObj != NULL)
{
hRes = aList[i].lpStream->Clone((IStream**)&(lpCd[nItems].pUnk));
if (FAILED(hRes))
break;
lpCd[nItems].dwCookie = (DWORD)(i+1);
nItems++;
}
}
if (FAILED(hRes))
{
for (i=0; i<nItems; i++)
((IStream*)(lpCd[nItems].pUnk))->Release();
delete lpCd;
delete pEnum;
return hRes;
}
}
for (i=0; i<nItems; i++)
{
cStream.Release();
cStream.Attach((IStream*)(lpCd[i].pUnk));
li.QuadPart = 0;
hRes = cStream->Seek(li, SEEK_SET, NULL);
if (SUCCEEDED(hRes))
{
p.Release();
hRes = ::CoUnmarshalInterface(cStream, IID_IDispatch, (void**)&p);
}
if (SUCCEEDED(hRes))
hRes = p->QueryInterface<IUnknown>(&(lpCd[i].pUnk));
if (FAILED(hRes))
{
for (k=0; k<nItems; k++)
{
if (k < i)
{
lpCd[k].pUnk->Release(); //pUnk are real IUnknown interfaces
}
else if (k > i)
{
((IStream*)(lpCd[k].pUnk))->Release(); //pUnk are IStream interfaces
}
}
return hRes;
}
}
//don't copy the data, but transfer ownership to it
pEnum->Init(lpCd, lpCd+nItems, NULL, AtlFlagTakeOwnership);
hRes = pEnum->_InternalQueryInterface(__uuidof(IEnumConnections), (void**)ppEnum);
if (FAILED(hRes))
delete pEnum;
return hRes;
};
HRESULT GetConnAt(__deref_out IDispatch **lplpDisp, __in int nIndex)
{
T* pT = static_cast<T*>(this);
TNktComPtr<IStream> cStream;
LARGE_INTEGER li;
HRESULT hRes;
NKT_ASSERT(lplpDisp != NULL);
*lplpDisp = NULL;
{
CNktAutoFastMutex cLock(&cMtx);
if (nIndex >= 0 && (SIZE_T)nIndex < aList.GetCount())
{
if (aList[nIndex].lpObj != NULL)
{
hRes = aList[nIndex].lpStream->Clone(&cStream);
if (SUCCEEDED(hRes))
{
li.QuadPart = 0;
hRes = cStream->Seek(li, SEEK_SET, NULL);
}
}
else
{
hRes = E_FAIL;
}
}
else
{
hRes = E_INVALIDARG;
}
}
if (SUCCEEDED(hRes))
hRes = ::CoUnmarshalInterface(cStream, IID_IDispatch, (void**)lplpDisp);
if (FAILED(hRes))
return (hRes == E_INVALIDARG) ? E_INVALIDARG : E_FAIL;
return S_OK;
};
//NOTE: May return DISP_E_MEMBERNOTFOUND
HRESULT FireCommon(__inout IDispatch *lpConnDisp, __in DISPID nDispId,
__in CComVariant aVarParams[], __in int aVarParamsCount)
{
DISPPARAMS sParams = { aVarParams, NULL, (UINT)aVarParamsCount, 0 };
CComVariant cVarResult;
HRESULT hRes;
UINT nArgErr;
try
{
hRes = lpConnDisp->Invoke(nDispId, IID_NULL, LOCALE_USER_DEFAULT, DISPATCH_METHOD, &sParams,
&cVarResult, NULL, &nArgErr);
}
catch (...)
{
hRes = E_FAIL;
}
return hRes;
};
private:
typedef struct {
IDispatch *lpObj;
IStream *lpStream;
} ITEM;
CNktFastMutex cMtx;
TNktArrayList4Structs<ITEM> aList;
};
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.