-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathoutputProcess.py
137 lines (126 loc) · 4.74 KB
/
outputProcess.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
#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Time : 19-4-23 上午9:40
# @Author : liyang
# @File : outputProcess.py
import pandas as pd
from tqdm import tqdm
# 在预测结果中寻找4所在位置。
def getFourIndex(arr):
indexs = []
for i in range(len(arr)):
if int(arr[i]) == 4:
indexs.append(i)
if len(indexs) != 2:
print(arr)
return indexs
# 将分词结果转化为1维，直接拉成一维
def convertTo1D(tokens):
res = []
for token in tokens:
res.extend(token)
return res
# 将预测结果转化为字符（存在'[UNK]'需要处理）
def getResultForEntity():
res = []
with open('data_dir/model/test_results.tsv', 'r') as f:
for line in f.read().strip().split('\n'):
example_out = line.strip().split('\t')
res.append(example_out)
print("results文件读取完成")
print(res[0])
tqdm.pandas(desc="getLabel:")
test_data = pd.read_json('data_dir/test_title.json', orient='records', lines=True)
print("json文件读取完成")
now_index = 0
def getLabel(example):
nonlocal now_index
sequences_tokens = example[0]
title_tokens = convertTo1D(example[1])
seq_cnt = len(sequences_tokens)
keywords = list()
for seq_index in range(seq_cnt):
seq_tokens = convertTo1D(sequences_tokens[seq_index])
seq_res = res[now_index+seq_index]
index4 = getFourIndex(seq_res)
seq_labels = seq_res[1: index4[0]]
title_labels = seq_res[index4[0]+1: index4[-1]]
i = 0
temp_str = ''
while i < len(seq_labels):
if seq_labels[i] == '1':
temp_str += seq_tokens[i].replace('##', '')
elif seq_labels[i] == '2':
if temp_str != '':
temp_str += seq_tokens[i].replace('##', '')
else:
if temp_str != '':
keywords.append(temp_str)
temp_str = ''
i += 1
i = 0
temp_str = ''
assert len(title_tokens) >= len(title_labels)
while i < len(title_labels):
if title_labels[i] == '1':
temp_str += title_tokens[i].replace('##', '')
elif title_labels[i] == '2':
if len(temp_str) != 0:
temp_str += title_tokens[i].replace('##', '')
else:
if temp_str != '':
keywords.append(temp_str)
temp_str = ''
i += 1
if "" in keywords:
keywords.remove('')
keywords = list(set(keywords))
level = set()
for i in range(len(keywords)):
for j in range(i+1, len(keywords)):
if keywords[i] in keywords[j]:
level.add(keywords[i])
if keywords[j] in keywords[i]:
level.add(keywords[j])
for item in list(level):
keywords.remove(item)
now_index += seq_cnt
return keywords
test_data['entity'] = test_data['example'].progress_apply(getLabel)
test_data[['newsId', 'entity']].to_json("/mnt/souhu/code/ly/data_dir/entity_output.json",
orient='records', lines=True, force_ascii=False)
'''
处理emotion输出
'''
def getResultForEmotion():
res = []
with open('data_dir/emotion_model/test_results.tsv', 'r') as f:
for line in f.read().strip().split('\n'):
example_out = line.strip()
res.append(example_out)
print("results文件读取完成")
tqdm.pandas(desc="getLabel:")
test_data = pd.read_json('data_dir/test_emotion.json', orient='records', lines=True)
print("json文件读取完成")
now_index = 0
def getEmotion(example):
emo_map = {'0': 'POS', '1': 'NORM', '2': 'NEG'}
nonlocal now_index
emotions = []
for i in range(len(example)):
emo = res[now_index]
emotions.append(emo_map[emo])
now_index += 1
return emotions
tqdm.pandas(desc="getLabel:")
test_data['emotion'] = test_data['emotionExamples'].progress_apply(getEmotion)
def getEntity(example):
entities = []
for i in range(len(example)):
entities.append(example[i][0])
return entities
test_data['entity'] = test_data['emotionExamples'].progress_apply(getEntity)
test_data[['newsId', 'entity', 'emotion']].to_json("/mnt/souhu/code/ly/data_dir/emotion_output.json",
orient='records', lines=True, force_ascii=False)
if __name__ == '__main__':
getResultForEmotion()