-
Notifications
You must be signed in to change notification settings - Fork 2
/
judger.py
157 lines (138 loc) · 5.35 KB
/
judger.py
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
import json
class Judger:
# Initialize Judger, with the path of tag list
def __init__(self, tag_path):
self.tag_dic = {}
f = open(tag_path, "r", encoding='utf-8')
self.task_cnt = 0
for line in f:
# print(line)
self.task_cnt += 1
self.tag_dic[line[:-1]] = self.task_cnt
# print(self.tag_dic)
# Format the result generated by the Predictor class
@staticmethod
def format_result(result):
rex = {"tags": []}
res_art = []
for x in result["tags"]:
if not (x is None):
res_art.append(int(x))
rex["tags"] = res_art
return rex
# Gen new results according to the truth and users output
def gen_new_result(self, result, truth, label):
s1 = set()
for tag in label:
s1.add(self.tag_dic.setdefault(tag.replace(' ', ''), None))
s2 = set()
for name in truth:
s2.add(self.tag_dic.setdefault(name.replace(' ', ''), None))
for a in range(0, self.task_cnt):
in1 = (a + 1) in s1
in2 = (a + 1) in s2
if in1:
if in2:
result[0][a]["TP"] += 1
else:
result[0][a]["FP"] += 1
else:
if in2:
result[0][a]["FN"] += 1
else:
result[0][a]["TN"] += 1
return result
# Calculate precision, recall and f1 value
# According to https://github.com/dice-group/gerbil/wiki/Precision,-Recall-and-F1-measure
@staticmethod
def get_value(res):
if res["TP"] == 0:
if res["FP"] == 0 and res["FN"] == 0:
precision = 1.0
recall = 1.0
f1 = 1.0
else:
precision = 0.0
recall = 0.0
f1 = 0.0
else:
precision = 1.0 * res["TP"] / (res["TP"] + res["FP"])
recall = 1.0 * res["TP"] / (res["TP"] + res["FN"])
f1 = 2 * precision * recall / (precision + recall)
return precision, recall, f1
# Generate score
def gen_score(self, arr):
sumf = 0
y = {"TP": 0, "FP": 0, "FN": 0, "TN": 0}
for x in arr[0]:
p, r, f = self.get_value(x)
sumf += f
for z in x.keys():
y[z] += x[z]
_, __, f_ = self.get_value(y)
macro_f = sumf * 1.0 / len(arr[0])
micro_f = f_
return {"macro" : macro_f, "micro" : micro_f}
# Test with ground truth path and the user's output path
def test(self, truth_path, output_path):
cnt = 0
result = [[]]
for a in range(0, self.task_cnt):
result[0].append({"TP": 0, "FP": 0, "TN": 0, "FN": 0})
# with open(truth_path, "r", encoding='utf-8') as inf, open(output_path, "r", encoding='utf-8') as ouf:
ground_doc_dict = {}
with open(truth_path, "r", encoding='utf-8') as inf:
for line in inf:
ground_doc = json.loads(line)
ah = ground_doc[0]['ah']
ground_doc_dict[ah] = ground_doc
with open(output_path, "r", encoding='utf-8') as inf:
for line in inf:
user_doc = json.loads(line)
ah = user_doc[0]['ah']
if ah in ground_doc_dict:
ground_doc = ground_doc_dict[ah]
else:
print("WARNING: ah", ah, "is not in ground truth file")
continue
for ind in range(len(ground_doc)):
ground_truth = ground_doc[ind]['label']
try:
user_output = user_doc[ind]['label']
except:
print(user_doc[ind])
cnt += 1
result = self.gen_new_result(result, ground_truth, user_output)
return result
# Generatue final_score
def get_score(truth_path_labor, output_path_labor, tag_path_labor):
def ret_operate_helper(info, ret, tag, score):
if tag == "total":
for flag in ["macro", "micro"]:
name = "%s_%s_f1" % (tag, flag)
info += "%14s : %10f\n" % (name, score[flag])
ret.append((name, score[flag]))
else:
flag = "macro"
name = "%s_f1" % (tag)
info += "%14s : %10f\n" % (name, score[flag])
ret.append((name, score[flag]))
return info, ret
judger_labor = Judger(tag_path=tag_path_labor)
reslt_labor = judger_labor.test(truth_path=truth_path_labor,
output_path=output_path_labor)
tag_dic = list(judger_labor.tag_dic)
ret = []
ret_str = ""
for idx, d in enumerate(reslt_labor[0]):
f1 = judger_labor.gen_score([[d]])
ret_str, ret = ret_operate_helper(ret_str, ret, tag_dic[idx], f1)
total_f1 = judger_labor.gen_score(reslt_labor)
ret_str, ret = ret_operate_helper(ret_str, ret, "total", total_f1)
return ret, ret_str[:-1]
if __name__ == '__main__':
final_score = get_score(truth_path_labor='test_data/10_new.json',
output_path_labor='test_data/abl_predict_0.json',
tag_path_labor='test_data/tags_for_test.txt')
print(final_score)
print(final_score[1])