File size: 6,905 Bytes
2a26d3b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 |
"""Information Retrieval metrics
Useful Resources:
http://www.cs.utexas.edu/~mooney/ir-course/slides/Evaluation.ppt
http://www.nii.ac.jp/TechReports/05-014E.pdf
http://www.stanford.edu/class/cs276/handouts/EvaluationNew-handout-6-per.pdf
http://hal.archives-ouvertes.fr/docs/00/72/67/60/PDF/07-busa-fekete.pdf
Learning to Rank for Information Retrieval (Tie-Yan Liu)
"""
import numpy as np
import pdb
import os
import pickle
def mean_reciprocal_rank(rs):
"""Score is reciprocal of the rank of the first relevant item
First element is 'rank 1'. Relevance is binary (nonzero is relevant).
Example from http://en.wikipedia.org/wiki/Mean_reciprocal_rank
>>> rs = [[0, 0, 1], [0, 1, 0], [1, 0, 0]]
>>> mean_reciprocal_rank(rs)
0.61111111111111105
>>> rs = np.array([[0, 0, 0], [0, 1, 0], [1, 0, 0]])
>>> mean_reciprocal_rank(rs)
0.5
>>> rs = [[0, 0, 0, 1], [1, 0, 0], [1, 0, 0]]
>>> mean_reciprocal_rank(rs)
0.75
Args:
rs: Iterator of relevance scores (list or numpy) in rank order
(first element is the first item)
Returns:
Mean reciprocal rank
"""
rs = (np.asarray(r).nonzero()[0] for r in rs)
return np.mean([1. / (r[0] + 1) if r.size else 0. for r in rs])
def r_precision(r):
"""Score is precision after all relevant documents have been retrieved
Relevance is binary (nonzero is relevant).
>>> r = [0, 0, 1]
>>> r_precision(r)
0.33333333333333331
>>> r = [0, 1, 0]
>>> r_precision(r)
0.5
>>> r = [1, 0, 0]
>>> r_precision(r)
1.0
Args:
r: Relevance scores (list or numpy) in rank order
(first element is the first item)
Returns:
R Precision
"""
r = np.asarray(r) != 0
z = r.nonzero()[0]
if not z.size:
return 0.
return np.mean(r[:z[-1] + 1])
def precision_at_k(r, k):
"""Score is precision @ k
Relevance is binary (nonzero is relevant).
>>> r = [0, 0, 1]
>>> precision_at_k(r, 1)
0.0
>>> precision_at_k(r, 2)
0.0
>>> precision_at_k(r, 3)
0.33333333333333331
>>> precision_at_k(r, 4)
Traceback (most recent call last):
File "<stdin>", line 1, in ?
ValueError: Relevance score length < k
Args:
r: Relevance scores (list or numpy) in rank order
(first element is the first item)
Returns:
Precision @ k
Raises:
ValueError: len(r) must be >= k
"""
assert k >= 1
r = np.asarray(r)[:k] != 0
if r.size != k:
raise ValueError('Relevance score length < k')
return np.mean(r)
def average_precision(r):
"""Score is average precision (area under PR curve)
Relevance is binary (nonzero is relevant).
>>> r = [1, 1, 0, 1, 0, 1, 0, 0, 0, 1]
>>> delta_r = 1. / sum(r)
>>> sum([sum(r[:x + 1]) / (x + 1.) * delta_r for x, y in enumerate(r) if y])
0.7833333333333333
>>> average_precision(r)
0.78333333333333333
Args:
r: Relevance scores (list or numpy) in rank order
(first element is the first item)
Returns:
Average precision
"""
r = np.asarray(r) != 0
out = [precision_at_k(r, k + 1) for k in range(r.size) if r[k]]
if not out:
return 0.
return np.mean(out)
def mean_average_precision(rs):
"""Score is mean average precision
Relevance is binary (nonzero is relevant).
>>> rs = [[1, 1, 0, 1, 0, 1, 0, 0, 0, 1]]
>>> mean_average_precision(rs)
0.78333333333333333
>>> rs = [[1, 1, 0, 1, 0, 1, 0, 0, 0, 1], [0]]
>>> mean_average_precision(rs)
0.39166666666666666
Args:
rs: Iterator of relevance scores (list or numpy) in rank order
(first element is the first item)
Returns:
Mean average precision
"""
return np.mean([average_precision(r) for r in rs])
def row_pop_average_precision(r, target):
"""Score is average precision (area under PR curve)
Relevance is binary (nonzero is relevant).
>>> r = [1, 1, 0, 1, 0, 1, 0, 0, 0, 1]
>>> delta_r = 1. / sum(r)
>>> sum([sum(r[:x + 1]) / (x + 1.) * delta_r for x, y in enumerate(r) if y])
0.7833333333333333
>>> average_precision(r)
0.78333333333333333
Args:
r: Relevance scores (list or numpy) in rank order
(first element is the first item)
Returns:
Average precision
"""
r = np.asarray(r) != 0
out = [precision_at_k(r, k + 1) for k in range(r.size) if r[k]]
if len(out) < len(target):
out += [0] * (len(target) - len(out))
if not out:
return 0.
return np.mean(out)
def dcg_at_k(r, k, method=0):
"""Score is discounted cumulative gain (dcg)
Relevance is positive real values. Can use binary
as the previous methods.
Example from
http://www.stanford.edu/class/cs276/handouts/EvaluationNew-handout-6-per.pdf
>>> r = [3, 2, 3, 0, 0, 1, 2, 2, 3, 0]
>>> dcg_at_k(r, 1)
3.0
>>> dcg_at_k(r, 1, method=1)
3.0
>>> dcg_at_k(r, 2)
5.0
>>> dcg_at_k(r, 2, method=1)
4.2618595071429155
>>> dcg_at_k(r, 10)
9.6051177391888114
>>> dcg_at_k(r, 11)
9.6051177391888114
Args:
r: Relevance scores (list or numpy) in rank order
(first element is the first item)
k: Number of results to consider
method: If 0 then weights are [1.0, 1.0, 0.6309, 0.5, 0.4307, ...]
If 1 then weights are [1.0, 0.6309, 0.5, 0.4307, ...]
Returns:
Discounted cumulative gain
"""
r = np.asfarray(r)[:k]
if r.size:
if method == 0:
return r[0] + np.sum(r[1:] / np.log2(np.arange(2, r.size + 1)))
elif method == 1:
return np.sum(r / np.log2(np.arange(2, r.size + 2)))
else:
raise ValueError('method must be 0 or 1.')
return 0.
def ndcg_at_k(r, k, method=0):
"""Score is normalized discounted cumulative gain (ndcg)
Relevance is positive real values. Can use binary
as the previous methods.
Example from
http://www.stanford.edu/class/cs276/handouts/EvaluationNew-handout-6-per.pdf
>>> r = [3, 2, 3, 0, 0, 1, 2, 2, 3, 0]
>>> ndcg_at_k(r, 1)
1.0
>>> r = [2, 1, 2, 0]
>>> ndcg_at_k(r, 4)
0.9203032077642922
>>> ndcg_at_k(r, 4, method=1)
0.96519546960144276
>>> ndcg_at_k([0], 1)
0.0
>>> ndcg_at_k([1], 2)
1.0
Args:
r: Relevance scores (list or numpy) in rank order
(first element is the first item)
k: Number of results to consider
method: If 0 then weights are [1.0, 1.0, 0.6309, 0.5, 0.4307, ...]
If 1 then weights are [1.0, 0.6309, 0.5, 0.4307, ...]
Returns:
Normalized discounted cumulative gain
"""
dcg_max = dcg_at_k(sorted(r, reverse=True), k, method)
if not dcg_max:
return 0.
return dcg_at_k(r, k, method) / dcg_max
|