prompt
large_stringlengths 72
9.34k
| completion
large_stringlengths 0
7.61k
|
|---|---|
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
<|fim_middle|>
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2 |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
<|fim_middle|>
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
<|fim_middle|>
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
<|fim_middle|>
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop; |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
<|fim_middle|>
<｜fim▁end｜> | popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
<|fim_middle|>
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | return 0 |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
<|fim_middle|>
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | return 1 |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
<|fim_middle|>
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | return 2 |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
<|fim_middle|>
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | lineNum += 1
continue |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
<|fim_middle|>
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | pop = 1 |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
<|fim_middle|>
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | pop = 10 |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
<|fim_middle|>
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | pop = 100 |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def <|fim_middle|>(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | getActionScore |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def <|fim_middle|>(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | compute_interaction |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def <|fim_middle|>(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | compute_user_history_interaction |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def <|fim_middle|>(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def compute_user_popularity(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | get_action_weight |
<|file_name|>compute_user_popularity.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
# compute the times of action(rec|click|msg) for each user
from math import sqrt
def getActionScore(action):
if action == "rec":
return 0
elif action == "click" :
return 1
else:
return 2
def compute_interaction(data):
interaction = {}
for line in data:
(userA,userB,times,action) = line.split(' ')
action = action[:-1]
key = userB + " " + action
interaction.setdefault(key, 0)
interaction[key] += 1
return interaction
def compute_user_history_interaction(trainFile):
records = []
lineList = []
lineNum = 1
result = []
lineList = [line for line in file(trainFile)]
for line in lineList:
if lineNum == 1: #ignore the title in first line
lineNum += 1
continue
records.append(line)
lineNum += 1
interaction = compute_interaction(records)
out = file('user_interaction.txt', 'w')
for (key, times) in interaction.items():
out.write('%s %d' % (key, times))
out.write('\n')
for (key, times) in interaction.items():
user, action = key.split(' ');
result.append((user, action, times))
return result
#get the weight for each type of action
def get_action_weight(action):
pop = 0;
if action == "rec":
pop = 1
elif action == "click":
pop = 10
elif action == "msg":
pop = 100
return pop;
#trainFile line like: [userA, userB, action_times, action_type(rec|click|msg)]
def <|fim_middle|>(trainFile, user_popularity_file):
popDict = {}
rankedscores = []
result = []
print "-----compute_user_history_interaction ... "
interaction = compute_user_history_interaction(trainFile)
print "-----compute_user_popularity ... "
for (user, action, times) in interaction[0:len(interaction)]:
popDict.setdefault(user, 0)
popDict[user] += get_action_weight(action) * times
ranked_popularity = [(popularity, user) for (user, popularity) in popDict.items()]
ranked_popularity.sort()
ranked_popularity.reverse()
print "-----ranking_user_popularity ... "
result = [(user, popularity) for (popularity, user) in ranked_popularity[0:len(ranked_popularity)]]
print "-----output user_popularity ... "
out = file(user_popularity_file, 'w')
for (user, popularity) in result[0:len(result)]:
out.write('%s %d\n' % (user, popularity))
print "-----Ending ... "
return result<｜fim▁end｜> | compute_user_popularity |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)<｜fim▁hole｜> gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()<｜fim▁end｜> | mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
<|fim_middle|>
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
<|fim_middle|>
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
<|fim_middle|>
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
<|fim_middle|>
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
<|fim_middle|>
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
<|fim_middle|>
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
<|fim_middle|>
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
<|fim_middle|>
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
<|fim_middle|>
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
<|fim_middle|>
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
<|fim_middle|>
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda() |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
<|fim_middle|>
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
<|fim_middle|>
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | torch.cuda.manual_seed_all(0) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
<|fim_middle|>
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | torch.set_rng_state(self.rng_state) |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
<|fim_middle|>
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | return |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
<|fim_middle|>
<｜fim▁end｜> | unittest.main() |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def <|fim_middle|>(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | make_data |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def <|fim_middle|>(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | __init__ |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def <|fim_middle|>(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | forward |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def <|fim_middle|>(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | setUp |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def <|fim_middle|>(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | tearDown |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def <|fim_middle|>(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | test_sgpr_mean_abs_error |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def <|fim_middle|>(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def test_sgpr_mean_abs_error_cuda(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | test_sgpr_fast_pred_var |
<|file_name|>test_sgpr_regression.py<|end_file_name|><｜fim▁begin｜>#!/usr/bin/env python3
import os
import random
import unittest
import warnings
from math import exp, pi
import gpytorch
import torch
from gpytorch.distributions import MultivariateNormal
from gpytorch.kernels import InducingPointKernel, RBFKernel, ScaleKernel
from gpytorch.likelihoods import GaussianLikelihood
from gpytorch.means import ConstantMean
from gpytorch.priors import SmoothedBoxPrior
from gpytorch.test.utils import least_used_cuda_device
from gpytorch.utils.warnings import NumericalWarning
from torch import optim
# Simple training data: let's try to learn a sine function,
# but with SGPR
# let's use 100 training examples.
def make_data(cuda=False):
train_x = torch.linspace(0, 1, 100)
train_y = torch.sin(train_x * (2 * pi))
train_y.add_(torch.randn_like(train_y), alpha=1e-2)
test_x = torch.rand(51)
test_y = torch.sin(test_x * (2 * pi))
if cuda:
train_x = train_x.cuda()
train_y = train_y.cuda()
test_x = test_x.cuda()
test_y = test_y.cuda()
return train_x, train_y, test_x, test_y
class GPRegressionModel(gpytorch.models.ExactGP):
def __init__(self, train_x, train_y, likelihood):
super(GPRegressionModel, self).__init__(train_x, train_y, likelihood)
self.mean_module = ConstantMean(prior=SmoothedBoxPrior(-1e-5, 1e-5))
self.base_covar_module = ScaleKernel(RBFKernel(lengthscale_prior=SmoothedBoxPrior(exp(-5), exp(6), sigma=0.1)))
self.covar_module = InducingPointKernel(
self.base_covar_module, inducing_points=torch.linspace(0, 1, 32), likelihood=likelihood
)
def forward(self, x):
mean_x = self.mean_module(x)
covar_x = self.covar_module(x)
return MultivariateNormal(mean_x, covar_x)
class TestSGPRRegression(unittest.TestCase):
def setUp(self):
if os.getenv("UNLOCK_SEED") is None or os.getenv("UNLOCK_SEED").lower() == "false":
self.rng_state = torch.get_rng_state()
torch.manual_seed(0)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(0)
random.seed(0)
def tearDown(self):
if hasattr(self, "rng_state"):
torch.set_rng_state(self.rng_state)
def test_sgpr_mean_abs_error(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(30):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.05)
def test_sgpr_fast_pred_var(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
train_x, train_y, test_x, test_y = make_data()
likelihood = GaussianLikelihood()
gp_model = GPRegressionModel(train_x, train_y, likelihood)
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
for _ in range(50):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
with gpytorch.settings.max_preconditioner_size(5), gpytorch.settings.max_cg_iterations(50):
with gpytorch.settings.fast_pred_var(True):
fast_var = gp_model(test_x).variance
fast_var_cache = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - fast_var).abs()), 1e-3)
with gpytorch.settings.fast_pred_var(False):
slow_var = gp_model(test_x).variance
self.assertLess(torch.max((fast_var_cache - slow_var).abs()), 1e-3)
def <|fim_middle|>(self):
# Suppress numerical warnings
warnings.simplefilter("ignore", NumericalWarning)
if not torch.cuda.is_available():
return
with least_used_cuda_device():
train_x, train_y, test_x, test_y = make_data(cuda=True)
likelihood = GaussianLikelihood().cuda()
gp_model = GPRegressionModel(train_x, train_y, likelihood).cuda()
mll = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, gp_model)
# Optimize the model
gp_model.train()
likelihood.train()
optimizer = optim.Adam(gp_model.parameters(), lr=0.1)
optimizer.n_iter = 0
for _ in range(25):
optimizer.zero_grad()
output = gp_model(train_x)
loss = -mll(output, train_y)
loss.backward()
optimizer.n_iter += 1
optimizer.step()
for param in gp_model.parameters():
self.assertTrue(param.grad is not None)
self.assertGreater(param.grad.norm().item(), 0)
# Test the model
gp_model.eval()
likelihood.eval()
test_preds = likelihood(gp_model(test_x)).mean
mean_abs_error = torch.mean(torch.abs(test_y - test_preds))
self.assertLess(mean_abs_error.squeeze().item(), 0.02)
if __name__ == "__main__":
unittest.main()
<｜fim▁end｜> | test_sgpr_mean_abs_error_cuda |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project<｜fim▁hole｜>
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))<｜fim▁end｜> | import os
import shutil
import sys
import SocketServer |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
<|fim_middle|>
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | """Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH) |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
<|fim_middle|>
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | """Build local version of site"""
local('pelican -s pelicanconf.py') |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
<|fim_middle|>
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | """`build` with the delete switch"""
local('pelican -d -s pelicanconf.py') |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
<|fim_middle|>
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | """Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py') |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
<|fim_middle|>
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | """Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever() |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
<|fim_middle|>
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | allow_reuse_address = True |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
<|fim_middle|>
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | """`build`, then `serve`"""
build()
serve() |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
<|fim_middle|>
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | """Build production version of site"""
local('pelican -s publishconf.py') |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
<|fim_middle|>
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | """Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env)) |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
<|fim_middle|>
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | """Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
) |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
<|fim_middle|>
<｜fim▁end｜> | """Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env)) |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
<|fim_middle|>
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH) |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def <|fim_middle|>():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | clean |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def <|fim_middle|>():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | build |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def <|fim_middle|>():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | rebuild |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def <|fim_middle|>():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | regenerate |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def <|fim_middle|>():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | serve |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def <|fim_middle|>():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | reserve |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def <|fim_middle|>():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | preview |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def <|fim_middle|>():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | cf_upload |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def <|fim_middle|>():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def gh_pages():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | publish |
<|file_name|>fabfile.py<|end_file_name|><｜fim▁begin｜>from fabric.api import *
import fabric.contrib.project as project
import os
import shutil
import sys
import SocketServer
from pelican.server import ComplexHTTPRequestHandler
# Local path configuration (can be absolute or relative to fabfile)
env.deploy_path = 'output'
DEPLOY_PATH = env.deploy_path
# Remote server configuration
production = 'root@localhost:22'
dest_path = '/var/www'
# Rackspace Cloud Files configuration settings
env.cloudfiles_username = 'my_rackspace_username'
env.cloudfiles_api_key = 'my_rackspace_api_key'
env.cloudfiles_container = 'my_cloudfiles_container'
# Github Pages configuration
env.github_pages_branch = "master"
# Port for `serve`
PORT = 8000
def clean():
"""Remove generated files"""
if os.path.isdir(DEPLOY_PATH):
shutil.rmtree(DEPLOY_PATH)
os.makedirs(DEPLOY_PATH)
def build():
"""Build local version of site"""
local('pelican -s pelicanconf.py')
def rebuild():
"""`build` with the delete switch"""
local('pelican -d -s pelicanconf.py')
def regenerate():
"""Automatically regenerate site upon file modification"""
local('pelican -r -s pelicanconf.py')
def serve():
"""Serve site at http://localhost:8000/"""
os.chdir(env.deploy_path)
class AddressReuseTCPServer(SocketServer.TCPServer):
allow_reuse_address = True
server = AddressReuseTCPServer(('', PORT), ComplexHTTPRequestHandler)
sys.stderr.write('Serving on port {0} ...\n'.format(PORT))
server.serve_forever()
def reserve():
"""`build`, then `serve`"""
build()
serve()
def preview():
"""Build production version of site"""
local('pelican -s publishconf.py')
def cf_upload():
"""Publish to Rackspace Cloud Files"""
rebuild()
with lcd(DEPLOY_PATH):
local('swift -v -A https://auth.api.rackspacecloud.com/v1.0 '
'-U {cloudfiles_username} '
'-K {cloudfiles_api_key} '
'upload -c {cloudfiles_container} .'.format(**env))
@hosts(production)
def publish():
"""Publish to production via rsync"""
local('pelican -s publishconf.py')
project.rsync_project(
remote_dir=dest_path,
exclude=".DS_Store",
local_dir=DEPLOY_PATH.rstrip('/') + '/',
delete=True,
extra_opts='-c',
)
def <|fim_middle|>():
"""Publish to GitHub Pages"""
rebuild()
local("ghp-import -b {github_pages_branch} {deploy_path} -p".format(**env))
<｜fim▁end｜> | gh_pages |
<|file_name|>parts_descriptor_test.py<|end_file_name|><｜fim▁begin｜>import amitgroup as ag
import numpy as np
<｜fim▁hole｜>
pd = ag.features.PartsDescriptor((5, 5), 20, patch_frame=1, edges_threshold=5, samples_per_image=10)
# Use only 100 of the digits
pd.train_from_images(data)
# Save the model to a file.
#pd.save('parts_model.npy')
# You can then load it again by
#pd = ag.features.PartsDescriptor.load(filename)
# Then you can extract features by
#features = pd.extract_features(image)
# Visualize the parts
ag.plot.images(pd.visparts)<｜fim▁end｜> | ag.set_verbose(True)
# This requires you to have the MNIST data set.
data, digits = ag.io.load_mnist('training', selection=slice(0, 100)) |
<|file_name|>util.py<|end_file_name|><｜fim▁begin｜># proxy module<｜fim▁hole｜><｜fim▁end｜> | from apptools.logger.util import * |
<|file_name|>environmental_probe.py<|end_file_name|><｜fim▁begin｜># ===============================================================================
# Copyright 2014 Jake Ross
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ===============================================================================
# ============= enthought library imports =======================
# ============= standard library imports ========================
# ============= local library imports ==========================
from __future__ import absolute_import
from pychron.hardware.core.core_device import CoreDevice
class TempHumMicroServer(CoreDevice):
"""
http://www.omega.com/Manuals/manualpdf/M3861.pdf
iServer MicroServer
<｜fim▁hole｜> scan_func = 'read_temperature'
def read_temperature(self, **kw):
v = self.ask('*SRTF', timeout=1.0, **kw)
return self._parse_response(v)
def read_humidity(self, **kw):
v = self.ask('*SRH', timeout=1.0, **kw)
return self._parse_response(v)
def _parse_response(self, v):
try:
return float(v)
except (AttributeError, ValueError, TypeError):
return self.get_random_value()
# ============= EOF =============================================<｜fim▁end｜> | tested with iTHX-W
"""
|
<|file_name|>environmental_probe.py<|end_file_name|><｜fim▁begin｜># ===============================================================================
# Copyright 2014 Jake Ross
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ===============================================================================
# ============= enthought library imports =======================
# ============= standard library imports ========================
# ============= local library imports ==========================
from __future__ import absolute_import
from pychron.hardware.core.core_device import CoreDevice
class TempHumMicroServer(CoreDevice):
<|fim_middle|>
# ============= EOF =============================================
<｜fim▁end｜> | """
http://www.omega.com/Manuals/manualpdf/M3861.pdf
iServer MicroServer
tested with iTHX-W
"""
scan_func = 'read_temperature'
def read_temperature(self, **kw):
v = self.ask('*SRTF', timeout=1.0, **kw)
return self._parse_response(v)
def read_humidity(self, **kw):
v = self.ask('*SRH', timeout=1.0, **kw)
return self._parse_response(v)
def _parse_response(self, v):
try:
return float(v)
except (AttributeError, ValueError, TypeError):
return self.get_random_value() |
<|file_name|>environmental_probe.py<|end_file_name|><｜fim▁begin｜># ===============================================================================
# Copyright 2014 Jake Ross
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ===============================================================================
# ============= enthought library imports =======================
# ============= standard library imports ========================
# ============= local library imports ==========================
from __future__ import absolute_import
from pychron.hardware.core.core_device import CoreDevice
class TempHumMicroServer(CoreDevice):
"""
http://www.omega.com/Manuals/manualpdf/M3861.pdf
iServer MicroServer
tested with iTHX-W
"""
scan_func = 'read_temperature'
def read_temperature(self, **kw):
<|fim_middle|>
def read_humidity(self, **kw):
v = self.ask('*SRH', timeout=1.0, **kw)
return self._parse_response(v)
def _parse_response(self, v):
try:
return float(v)
except (AttributeError, ValueError, TypeError):
return self.get_random_value()
# ============= EOF =============================================
<｜fim▁end｜> | v = self.ask('*SRTF', timeout=1.0, **kw)
return self._parse_response(v) |
<|file_name|>environmental_probe.py<|end_file_name|><｜fim▁begin｜># ===============================================================================
# Copyright 2014 Jake Ross
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ===============================================================================
# ============= enthought library imports =======================
# ============= standard library imports ========================
# ============= local library imports ==========================
from __future__ import absolute_import
from pychron.hardware.core.core_device import CoreDevice
class TempHumMicroServer(CoreDevice):
"""
http://www.omega.com/Manuals/manualpdf/M3861.pdf
iServer MicroServer
tested with iTHX-W
"""
scan_func = 'read_temperature'
def read_temperature(self, **kw):
v = self.ask('*SRTF', timeout=1.0, **kw)
return self._parse_response(v)
def read_humidity(self, **kw):
<|fim_middle|>
def _parse_response(self, v):
try:
return float(v)
except (AttributeError, ValueError, TypeError):
return self.get_random_value()
# ============= EOF =============================================
<｜fim▁end｜> | v = self.ask('*SRH', timeout=1.0, **kw)
return self._parse_response(v) |
<|file_name|>environmental_probe.py<|end_file_name|><｜fim▁begin｜># ===============================================================================
# Copyright 2014 Jake Ross
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ===============================================================================
# ============= enthought library imports =======================
# ============= standard library imports ========================
# ============= local library imports ==========================
from __future__ import absolute_import
from pychron.hardware.core.core_device import CoreDevice
class TempHumMicroServer(CoreDevice):
"""
http://www.omega.com/Manuals/manualpdf/M3861.pdf
iServer MicroServer
tested with iTHX-W
"""
scan_func = 'read_temperature'
def read_temperature(self, **kw):
v = self.ask('*SRTF', timeout=1.0, **kw)
return self._parse_response(v)
def read_humidity(self, **kw):
v = self.ask('*SRH', timeout=1.0, **kw)
return self._parse_response(v)
def _parse_response(self, v):
<|fim_middle|>
# ============= EOF =============================================
<｜fim▁end｜> | try:
return float(v)
except (AttributeError, ValueError, TypeError):
return self.get_random_value() |
<|file_name|>environmental_probe.py<|end_file_name|><｜fim▁begin｜># ===============================================================================
# Copyright 2014 Jake Ross
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ===============================================================================
# ============= enthought library imports =======================
# ============= standard library imports ========================
# ============= local library imports ==========================
from __future__ import absolute_import
from pychron.hardware.core.core_device import CoreDevice
class TempHumMicroServer(CoreDevice):
"""
http://www.omega.com/Manuals/manualpdf/M3861.pdf
iServer MicroServer
tested with iTHX-W
"""
scan_func = 'read_temperature'
def <|fim_middle|>(self, **kw):
v = self.ask('*SRTF', timeout=1.0, **kw)
return self._parse_response(v)
def read_humidity(self, **kw):
v = self.ask('*SRH', timeout=1.0, **kw)
return self._parse_response(v)
def _parse_response(self, v):
try:
return float(v)
except (AttributeError, ValueError, TypeError):
return self.get_random_value()
# ============= EOF =============================================
<｜fim▁end｜> | read_temperature |
<|file_name|>environmental_probe.py<|end_file_name|><｜fim▁begin｜># ===============================================================================
# Copyright 2014 Jake Ross
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ===============================================================================
# ============= enthought library imports =======================
# ============= standard library imports ========================
# ============= local library imports ==========================
from __future__ import absolute_import
from pychron.hardware.core.core_device import CoreDevice
class TempHumMicroServer(CoreDevice):
"""
http://www.omega.com/Manuals/manualpdf/M3861.pdf
iServer MicroServer
tested with iTHX-W
"""
scan_func = 'read_temperature'
def read_temperature(self, **kw):
v = self.ask('*SRTF', timeout=1.0, **kw)
return self._parse_response(v)
def <|fim_middle|>(self, **kw):
v = self.ask('*SRH', timeout=1.0, **kw)
return self._parse_response(v)
def _parse_response(self, v):
try:
return float(v)
except (AttributeError, ValueError, TypeError):
return self.get_random_value()
# ============= EOF =============================================
<｜fim▁end｜> | read_humidity |
<|file_name|>environmental_probe.py<|end_file_name|><｜fim▁begin｜># ===============================================================================
# Copyright 2014 Jake Ross
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ===============================================================================
# ============= enthought library imports =======================
# ============= standard library imports ========================
# ============= local library imports ==========================
from __future__ import absolute_import
from pychron.hardware.core.core_device import CoreDevice
class TempHumMicroServer(CoreDevice):
"""
http://www.omega.com/Manuals/manualpdf/M3861.pdf
iServer MicroServer
tested with iTHX-W
"""
scan_func = 'read_temperature'
def read_temperature(self, **kw):
v = self.ask('*SRTF', timeout=1.0, **kw)
return self._parse_response(v)
def read_humidity(self, **kw):
v = self.ask('*SRH', timeout=1.0, **kw)
return self._parse_response(v)
def <|fim_middle|>(self, v):
try:
return float(v)
except (AttributeError, ValueError, TypeError):
return self.get_random_value()
# ============= EOF =============================================
<｜fim▁end｜> | _parse_response |
<|file_name|>testroute.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
from .base import BaseHandler<｜fim▁hole｜>class TestRoute(BaseHandler):
def get(self, file):
return self.render(str(file) + '.jade', show_h1=1)<｜fim▁end｜> | |
<|file_name|>testroute.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
from .base import BaseHandler
class TestRoute(BaseHandler):
<|fim_middle|>
<｜fim▁end｜> | def get(self, file):
return self.render(str(file) + '.jade', show_h1=1) |
<|file_name|>testroute.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
from .base import BaseHandler
class TestRoute(BaseHandler):
def get(self, file):
<|fim_middle|>
<｜fim▁end｜> | return self.render(str(file) + '.jade', show_h1=1) |
<|file_name|>testroute.py<|end_file_name|><｜fim▁begin｜># -*- coding: utf-8 -*-
from .base import BaseHandler
class TestRoute(BaseHandler):
def <|fim_middle|>(self, file):
return self.render(str(file) + '.jade', show_h1=1)
<｜fim▁end｜> | get |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
<｜fim▁hole｜><｜fim▁end｜> | print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
<|fim_middle|>
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
<|fim_middle|>
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids))) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
<|fim_middle|>
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
<|fim_middle|>
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | return self.site.storage.getSize(key) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
<|fim_middle|>
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
<|fim_middle|>
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
<|fim_middle|>
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
<|fim_middle|>
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
<|fim_middle|>
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
<|fim_middle|>
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
<|fim_middle|>
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
<|fim_middle|>
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
<|fim_middle|>
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
<|fim_middle|>
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | params["site_id"] = self.db_id
return self.db.execute(query, params) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
<|fim_middle|>
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key)) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
<|fim_middle|>
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack())) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
<|fim_middle|>
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key)) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
<|fim_middle|>
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | self.__delitem__(key) # File not exists anymore |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
<|fim_middle|>
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
self.cached_keys.append(key)
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False) |
<|file_name|>ContentDbDict.py<|end_file_name|><｜fim▁begin｜>import time
import os
import ContentDb
from Debug import Debug
from Config import config
class ContentDbDict(dict):
def __init__(self, site, *args, **kwargs):
s = time.time()
self.site = site
self.cached_keys = []
self.log = self.site.log
self.db = ContentDb.getContentDb()
self.db_id = self.db.needSite(site)
self.num_loaded = 0
super(ContentDbDict, self).__init__(self.db.loadDbDict(site)) # Load keys from database
self.log.debug("ContentDb init: %.3fs, found files: %s, sites: %s" % (time.time() - s, len(self), len(self.db.site_ids)))
def loadItem(self, key):
try:
self.num_loaded += 1
if self.num_loaded % 100 == 0:
if config.verbose:
self.log.debug("Loaded json: %s (latest: %s) called by: %s" % (self.num_loaded, key, Debug.formatStack()))
else:
self.log.debug("Loaded json: %s (latest: %s)" % (self.num_loaded, key))
content = self.site.storage.loadJson(key)
dict.__setitem__(self, key, content)
except IOError:
if dict.get(self, key):
self.__delitem__(key) # File not exists anymore
raise KeyError(key)
self.addCachedKey(key)
self.checkLimit()
return content
def getItemSize(self, key):
return self.site.storage.getSize(key)
# Only keep last 10 accessed json in memory
def checkLimit(self):
if len(self.cached_keys) > 10:
key_deleted = self.cached_keys.pop(0)
dict.__setitem__(self, key_deleted, False)
def addCachedKey(self, key):
if key not in self.cached_keys and key != "content.json" and len(key) > 40: # Always keep keys smaller than 40 char
<|fim_middle|>
def __getitem__(self, key):
val = dict.get(self, key)
if val: # Already loaded
return val
elif val is None: # Unknown key
raise KeyError(key)
elif val is False: # Loaded before, but purged from cache
return self.loadItem(key)
def __setitem__(self, key, val):
self.addCachedKey(key)
self.checkLimit()
size = self.getItemSize(key)
self.db.setContent(self.site, key, val, size)
dict.__setitem__(self, key, val)
def __delitem__(self, key):
self.db.deleteContent(self.site, key)
dict.__delitem__(self, key)
try:
self.cached_keys.remove(key)
except ValueError:
pass
def iteritems(self):
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
yield key, val
def items(self):
back = []
for key in dict.keys(self):
try:
val = self[key]
except Exception as err:
self.log.warning("Error loading %s: %s" % (key, err))
continue
back.append((key, val))
return back
def values(self):
back = []
for key, val in dict.iteritems(self):
if not val:
try:
val = self.loadItem(key)
except Exception:
continue
back.append(val)
return back
def get(self, key, default=None):
try:
return self.__getitem__(key)
except KeyError:
return default
except Exception as err:
self.site.bad_files[key] = self.site.bad_files.get(key, 1)
dict.__delitem__(self, key)
self.log.warning("Error loading %s: %s" % (key, err))
return default
def execute(self, query, params={}):
params["site_id"] = self.db_id
return self.db.execute(query, params)
if __name__ == "__main__":
import psutil
process = psutil.Process(os.getpid())
s_mem = process.memory_info()[0] / float(2 ** 20)
root = "data-live/1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27"
contents = ContentDbDict("1MaiL5gfBM1cyb4a8e3iiL8L5gXmoAJu27", root)
print "Init len", len(contents)
s = time.time()
for dir_name in os.listdir(root + "/data/users/")[0:8000]:
contents["data/users/%s/content.json" % dir_name]
print "Load: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key, val in contents.iteritems():
found += 1
assert key
assert val
print "Found:", found
print "Iteritem: %.3fs" % (time.time() - s)
s = time.time()
found = 0
for key in contents.keys():
found += 1
assert key in contents
print "In: %.3fs" % (time.time() - s)
print "Len:", len(contents.values()), len(contents.keys())
print "Mem: +", process.memory_info()[0] / float(2 ** 20) - s_mem
<｜fim▁end｜> | self.cached_keys.append(key) |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.