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
|
from tensorflow.python.keras.utils.generic_utils import to_list
from tensorflow.python.keras.utils import metrics_utils
from tensorflow.keras.initializers import zeros as zeros_initializer
from tensorflow.keras.metrics import Metric
import tensorflow.keras.backend as K
import tensorflow as tf
import numpy as np
class FScore(Metric):
'''Stateful Keras f-score metric'''
def __init__(self,
beta=1,
thresholds=None,
top_k=None,
class_id=None,
name='f1',
dtype=None,
argmax=False):
super(FScore, self).__init__(name=name, dtype=dtype)
self.beta = beta
self.init_thresholds = thresholds
self.top_k = top_k
self.class_id = class_id
self.argmax = argmax
default_threshold = 0.5 if top_k is None else metrics_utils.NEG_INF
self.thresholds = metrics_utils.parse_init_thresholds(
thresholds, default_threshold=default_threshold)
self.true_positives = self.add_weight(
'true_positives',
shape=(len(self.thresholds),),
initializer=zeros_initializer)
self.false_positives = self.add_weight(
'false_positives',
shape=(len(self.thresholds),),
initializer=zeros_initializer)
self.false_negatives = self.add_weight(
'false_negatives',
shape=(len(self.thresholds),),
initializer=zeros_initializer)
def update_state(self, y_true, y_pred, sample_weight=None):
'''Accumulates true positive and false positive statistics.'''
if self.argmax:
y_pred = K.argmax(y_pred)
return metrics_utils.update_confusion_matrix_variables(
{
metrics_utils.ConfusionMatrix.TRUE_POSITIVES: self.true_positives,
metrics_utils.ConfusionMatrix.FALSE_POSITIVES: self.false_positives,
metrics_utils.ConfusionMatrix.FALSE_NEGATIVES: self.false_negatives
},
y_true,
y_pred,
thresholds=self.thresholds,
top_k=self.top_k,
class_id=self.class_id,
sample_weight=sample_weight)
def result(self):
b2 = self.beta**2
q = (1+b2)*self.true_positives
d = (1+b2)*self.true_positives + b2*self.false_negatives + self.false_positives
result = tf.math.divide_no_nan(q, d)
return result[0] if len(self.thresholds) == 1 else result
def reset_state(self):
num_thresholds = len(to_list(self.thresholds))
K.batch_set_value(
[(v, np.zeros((num_thresholds,))) for v in self.variables])
def get_config(self):
config = {
'thresholds': self.init_thresholds,
'top_k': self.top_k,
'class_id': self.class_id,
'beta': self.beta,
}
base_config = super(FScore, self).get_config()
return dict(list(base_config.items()) + list(config.items()))
class MatthewsCorrCoef(Metric):
'''Stateful Keras Matthews correlation coefficient as a metric'''
def __init__(self,
thresholds=None,
top_k=None,
class_id=None,
name='mcc',
dtype=None,
argmax=False):
super(MatthewsCorrCoef, self).__init__(name=name, dtype=dtype)
self.init_thresholds = thresholds
self.top_k = top_k
self.class_id = class_id
self.argmax = argmax
default_threshold = 0.5 if top_k is None else metrics_utils.NEG_INF
self.thresholds = metrics_utils.parse_init_thresholds(
thresholds, default_threshold=default_threshold)
self.true_positives = self.add_weight(
'true_positives',
shape=(len(self.thresholds),),
initializer=zeros_initializer)
self.false_positives = self.add_weight(
'false_positives',
shape=(len(self.thresholds),),
initializer=zeros_initializer)
self.false_negatives = self.add_weight(
'false_negatives',
shape=(len(self.thresholds),),
initializer=zeros_initializer)
self.true_negatives = self.add_weight(
'true_negatives',
shape=(len(self.thresholds),),
initializer=zeros_initializer)
def update_state(self, y_true, y_pred, sample_weight=None):
'''Accumulates true positive and false positive statistics.'''
if self.argmax:
y_pred = K.argmax(y_pred)
return metrics_utils.update_confusion_matrix_variables(
{
metrics_utils.ConfusionMatrix.TRUE_POSITIVES: self.true_positives,
metrics_utils.ConfusionMatrix.FALSE_POSITIVES: self.false_positives,
metrics_utils.ConfusionMatrix.FALSE_NEGATIVES: self.false_negatives,
metrics_utils.ConfusionMatrix.TRUE_NEGATIVES: self.true_negatives
},
y_true,
y_pred,
thresholds=self.thresholds,
top_k=self.top_k,
class_id=self.class_id,
sample_weight=sample_weight)
def result(self):
N = (self.true_negatives + self.true_positives
+ self.false_negatives + self.false_positives)
S = (self.true_positives + self.false_negatives) / N
P = (self.true_positives + self.false_positives) / N
result = tf.math.divide_no_nan(self.true_positives/N - S*P,
tf.math.sqrt(P * S * (1-S) * (1-P)))
return result[0] if len(self.thresholds) == 1 else result
def reset_state(self):
num_thresholds = len(to_list(self.thresholds))
K.batch_set_value(
[(v, np.zeros((num_thresholds,))) for v in self.variables])
def get_config(self):
config = {
'thresholds': self.init_thresholds,
'top_k': self.top_k,
'class_id': self.class_id,
}
base_config = super(MatthewsCorrCoef, self).get_config()
return dict(list(base_config.items()) + list(config.items()))
|
