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
|
# -*- coding: utf-8 -*-
"""
>>> from pycm import *
>>> from pytest import warns
>>> cm_comp1 = ConfusionMatrix(matrix={0: {0: 2, 1: 50, 2: 6}, 1: {0: 5, 1: 50, 2: 3}, 2: {0: 1, 1: 7, 2: 50}})
>>> cm_comp2 = ConfusionMatrix(matrix={0: {0: 50, 1: 2, 2: 6}, 1: {0: 50, 1: 5, 2: 3}, 2: {0: 1, 1: 55, 2: 2}})
>>> cm_comp1 == cm_comp2
False
>>> cm_comp1 == 2
False
>>> cm_comp1_temp = ConfusionMatrix(matrix={0: {0: 2, 1: 50, 2: 6}, 1: {0: 5, 1: 50, 2: 3}, 2: {0: 1, 1: 7, 2: 50}})
>>> cm_comp1 == cm_comp1_temp
True
>>> compare_input = {"model1": cm_comp1, "model2": cm_comp2}
>>> compare_input_copy = {"model1": cm_comp1, "model2": cm_comp2}
>>> cp = Compare(compare_input)
>>> compare_input == compare_input_copy
True
>>> cp
pycm.Compare(classes: [0, 1, 2])
>>> cp.scores == {'model2': {'overall': 0.52857, 'class': 0.33611}, 'model1': {'overall': 0.58095, 'class': 0.50278}}
True
>>> cp.best
pycm.ConfusionMatrix(classes: [0, 1, 2])
>>> cp.best_name
'model1'
>>> print(cp)
Best : model1
<BLANKLINE>
Rank Name Class-Score Overall-Score
1 model1 0.50278 0.58095
2 model2 0.33611 0.52857
<BLANKLINE>
>>> cp.print_report()
Best : model1
<BLANKLINE>
Rank Name Class-Score Overall-Score
1 model1 0.50278 0.58095
2 model2 0.33611 0.52857
<BLANKLINE>
>>> class_weight = {0: 5, 1: 1, 2: 1}
>>> class_weight_copy = {0: 5, 1: 1, 2: 1}
>>> cp = Compare({"model1": cm_comp1, "model2": cm_comp2}, by_class=True, class_weight=class_weight)
>>> class_weight == class_weight_copy
True
>>> print(cp)
Best : model2
<BLANKLINE>
Rank Name Class-Score Overall-Score
1 model2 0.45357 0.52857
2 model1 0.34881 0.58095
<BLANKLINE>
>>> cp.best
pycm.ConfusionMatrix(classes: [0, 1, 2])
>>> cp.best_name
'model2'
>>> with warns(RuntimeWarning, match='Confusion matrices are too similar to identify a clear best option.'):
... cp2 = Compare({"model1": cm_comp1, "model2": cm_comp1})
>>> cp2.scores == {'model2': {'class': 0.50278, 'overall': 0.58095}, 'model1': {'class': 0.50278, 'overall': 0.58095}}
True
>>> cp2.best
>>> cp2.best_name
>>> cm1 = ConfusionMatrix(matrix={0: {0: 50, 1: 0, 2: 0}, 1: {0: 0, 1: 35, 2: 15}, 2: {0: 0, 1: 16, 2: 34}})
>>> cm2 = ConfusionMatrix(matrix={0: {0: 48, 1: 2, 2: 0}, 1: {0: 3, 1: 46, 2: 1}, 2: {0: 8, 1: 2, 2: 40}})
>>> cp3 = Compare({"cm1": cm1, "cm2": cm2})
>>> print(cp3)
Best : cm2
<BLANKLINE>
Rank Name Class-Score Overall-Score
1 cm2 0.70556 0.92381
2 cm1 0.55 0.75238
<BLANKLINE>
>>> cp4 = Compare({"cm1": cm1, "cm2": cm2}, class_weight={0: 0, 1: 1, 2: 1})
>>> cp4.class_weight == {0: 0, 1: 1, 2: 1}
True
>>> print(cp4)
Best : cm2
<BLANKLINE>
Rank Name Class-Score Overall-Score
1 cm2 0.825 0.92381
2 cm1 0.575 0.75238
<BLANKLINE>
>>> class_benchmark_weight = {"PLRI": 0, "NLRI": 0, "DPI": 0, "AUCI": 1, "MCCI": 0, "QI": 0}
>>> cp5 = Compare({"cm1": cm1, "cm2": cm2}, class_benchmark_weight=class_benchmark_weight)
>>> cp5.class_benchmark_weight == class_benchmark_weight
True
>>> print(cp5)
Best : cm2
<BLANKLINE>
Rank Name Class-Score Overall-Score
1 cm2 0.93333 0.92381
2 cm1 0.73333 0.75238
<BLANKLINE>
>>> overall_benchmark_weight = {"SOA1": 1, "SOA2": 0, "SOA3": 0, "SOA4": 0, "SOA5": 0, "SOA6": 1, "SOA7": 0, "SOA8": 0, "SOA9": 0, "SOA10": 0}
>>> cp6 = Compare({"cm1": cm1, "cm2": cm2}, class_benchmark_weight=class_benchmark_weight, overall_benchmark_weight=overall_benchmark_weight)
>>> cp6.overall_benchmark_weight == overall_benchmark_weight
True
>>> print(cp6)
Best : cm2
<BLANKLINE>
Rank Name Class-Score Overall-Score
1 cm2 0.93333 0.9
2 cm1 0.73333 0.71667
<BLANKLINE>
>>> with warns(RuntimeWarning, match='Confusion matrices are too similar to identify a clear best option.'):
... cp7 = Compare({"cm1": cm1, "cm2": cm2}, class_weight={0: 200, 1: 1, 2: 1})
>>> cp7.class_weight == {0: 200, 1: 1, 2: 1}
True
>>> print(cp7)
Best : None
<BLANKLINE>
Rank Name Class-Score Overall-Score
1 cm1 0.50074 0.75238
2 cm2 0.47021 0.92381
<BLANKLINE>
>>> cp7.best
>>> cp7.best_name
>>> y_actu = [2, 0, 2, 2, 0, 1, 1, 2, 2, 0, 1, 2]
>>> y_pred = [0, 0, 2, 1, 0, 2, 1, 0, 2, 0, 2, 2]
>>> cm = ConfusionMatrix(y_actu, y_pred)
>>> cm.relabel({0: "L1", 1: "L2", 2: "L3"})
>>> cm_null = ConfusionMatrix(matrix={0: {0: 0, 1: 0, 2: 0}, 1: {0: 0, 1: 0, 2: 0}, 2: {0: 0, 1: 0, 2: 0}})
>>> with warns(RuntimeWarning, match='Confusion matrices are too similar to identify a clear best option.'):
... cp8 = Compare({"cm1": cm, "cm2": cm}, class_weight={'L3': 6, 'L1': 3, 'L2': 3})
>>> with warns(RuntimeWarning, match='Invalid `class_weight` format; the result is for unweighted mode.'):
... cp9 = Compare({"cm1": cm1, "cm2": cm2}, class_weight={0: 0, 1: 0, 2: 0})
>>> class_benchmark_weight = {"PLRI": 0, "NLRI": 0, "DPI": 0, "AUCI": 0, "MCCI": 0, "QI": 0}
>>> with warns(RuntimeWarning, match='Invalid `class_benchmark_weight` format; the result is for unweighted mode.'):
... cp10 = Compare({"cm1": cm1, "cm2": cm2}, class_benchmark_weight=class_benchmark_weight)
>>> overall_benchmark_weight = {"SOA1": 0, "SOA2": 0, "SOA3": 0, "SOA4": 0, "SOA5": 0, "SOA6": 0, "SOA7": 0, "SOA8": 0, "SOA9": 0, "SOA10": 0}
>>> with warns(RuntimeWarning, match='Invalid `overall_benchmark_weight` format; the result is for unweighted mode.'):
... cp11 = Compare({"cm1": cm1, "cm2": cm2}, overall_benchmark_weight=overall_benchmark_weight)
>>> with warns(RuntimeWarning, match='Confusion matrices are too similar to identify a clear best option.'):
... cp12 = Compare({"cm1": cm_null, "cm2": cm_null})
"""
|
