-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathaita.py
executable file
·218 lines (191 loc) · 7.96 KB
/
aita.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
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
#! /usr/bin/python3
from WordAnalyser import WordAnalyser
from sklearn.ensemble import RandomForestRegressor, AdaBoostClassifier
import joblib
import re
import argparse
import matplotlib.pyplot as plt
import sklearn.svm as svm
# ! Trainer Definition
class Aita:
# ? answers_chunk -> 每个人的回答与评分整体内容
# ? answers_score -> 每个人的分数
# ? answers_main -> 每个人的回答的文字内容
# ? answers_len -> 每个人的回答的文字内容的长度
# ? answers_keywordcount -> 每个人的回答的文字内容中相关关键字出现次数
# ? classifier -> 训练得到的分类器
# ? 训练因子选择：[字数, 高频关键词统计...]
def __init__(self):
super().__init__()
word_analyser = WordAnalyser()
self.show_question_name()
self.answers_chunk = self.file_reader_withscore(
'./assets/marked_answers.txt')
self.answers_main = self.get_answers_main(self.answers_chunk)
self.top_words = word_analyser.keyword_analyse(self.answers_main)
self.classifier = None
def show_question_name(self):
# ? 显示问题
fp = open('./assets/question.txt')
print("针对的问题：《" + fp.readline()[:-1] + "》")
fp.close()
def file_reader_withscore(self, path):
# ? 读取答案数据，并将答案分组
train_file = open(path)
file_text = "".join(train_file.readlines())
regex_pattern = r'((\d{12}.*?#Total: *?\d\.*\d*/10))'
reg = re.compile(regex_pattern, re.DOTALL)
return reg.findall(file_text)
def file_reader(self, path):
# ? 读取答案数据，并将答案分组
train_file = open(path)
file_text = "".join(train_file.readlines())
regex_pattern = r'(\d{12}([^@#]*)?)'
reg = re.compile(regex_pattern, re.DOTALL)
return reg.findall(file_text)
def get_total_score(self, answers):
# ? 分析答案，获取评分
total_score = []
regex_pattern = r'#Total: *?(\d\.*\d*)/10'
reg = re.compile(regex_pattern, re.DOTALL)
for answer in answers:
total_score.append(float(reg.findall(answer[0])[0]))
return total_score
def get_answers_main(self, answers):
# ? 获取主要回答内容
answers_main = []
regex_pattern = r':\d\d\n([^@#]*)?'
reg = re.compile(regex_pattern, re.DOTALL)
for answer in answers:
answers_main.append(reg.findall(answer[0])[0])
return answers_main
def get_answers_len(self, answers):
# ? 获取回答长度
answers_len = []
for answer in answers:
answers_len.append(len(answer))
return answers_len
def get_keyword_counts(self, keywords, answers):
keyword_counts = []
for answer in answers:
per_keyword_counts = []
for i in keywords:
per_keyword_counts.append(answer.count(i[0]))
keyword_counts.append(per_keyword_counts)
return keyword_counts
def main_trainer(self):
self.answers_score = self.get_total_score(self.answers_chunk)
self.answers_len = self.get_answers_len(self.answers_chunk)
self.answers_keywordcount = self.get_keyword_counts(
self.top_words, self.answers_main)
train_data = []
for i in range(len(self.answers_main)):
per_line = [self.answers_len[i]]
per_line.extend(self.answers_keywordcount[i])
train_data.append(per_line)
print("------>>> 正在训练模型 <<<------")
clf = RandomForestRegressor(
n_estimators=100, # 子树数量
max_depth=8, # 最大深度
min_samples_leaf=6, # 叶节点最小样本数
max_features=7, # 节点分裂时参与判断的最大特征数
oob_score=True, # 使用袋外样品进行估算泛化精度
random_state=10 # 随机数种子
)
#clf= svm.NuSVR(nu = 0.5)
clf.fit(train_data, self.answers_score)
# ? 这一块是处理重要性的
impt = []
for i in range(len(self.top_words)):
impt.append([self.top_words[i][0], clf.feature_importances_[i]])
#print("(" + self.top_words[i][0] + ") \t重要性: " + str(clf.feature_importances_[i]))
impt.sort(key=lambda x: x[1], reverse=True)
for i in range(len(impt)):
print("(" + impt[i][0] + ") \t重要性: " + str(impt[i][1]))
# ? 保存模型
self.classifier = clf
print("----------------------------")
print("------>>> 训练模型结束 <<<------")
def predict(self, predict_answers):
#print("------>>> 预测结果如下 <<<------")
predict_keywords_count = self.get_keyword_counts(
self.top_words, [predict_answers])
predict_data = self.get_answers_len([predict_answers])
predict_data.extend(predict_keywords_count[0])
return self.classifier.predict([predict_data])[0]
# ! Trainer Definition End
def test():
# ? 测试模型，使用自带的回答
aita = Aita()
aita.classifier = joblib.load('Aita.model')
predict_chunk = aita.file_reader_withscore(
"assets/marked_answers2_simple.txt")
predict_true_score = aita.get_total_score(predict_chunk)
predict_main = aita.get_answers_main(predict_chunk)
mean_diff = 0
rss = 0
resi = []
for i in range(len(predict_main)):
predict_res = aita.predict(predict_main[i])
# print("\n----------------------------")
# print("预测结果: " + str(predict_res))
# print("实际结果: " + str(predict_true_score[i]))
# print("----------------------------")
resi.append((predict_true_score[i] - predict_res))
rss += (predict_true_score[i] - predict_res) ** 2
mean_diff += abs(predict_true_score[i] - predict_res)
mse = rss / len(predict_main)
print("平均误差: " + str(mean_diff / len(predict_main)))
print("MSE: " + str(mse))
# 下面都是画图的
'''
figure, axes=plt.subplots() #得到画板、轴
axes.boxplot(resi, patch_artist=True) #描点上色
plt.show()
plt.plot([x for x in range(len(predict_main))], resi, marker='o', color='red')
plt.rcParams['font.sans-serif'] = ['PingFang HK']
plt.rcParams['axes.unicode_minus'] = False
plt.title("预测结果",fontsize=14)
plt.xlabel("回答编号",fontsize=14)
plt.ylabel("误差^2", fontsize=14)
plt.gca().xaxis.set_major_locator(plt.MultipleLocator(1))
plt.gcf().set_size_inches(15, 7)
plt.show()
'''
def train():
# ? 训练模型
aita = Aita()
aita.main_trainer()
print("------>>> 正在保存模型 <<<------")
joblib.dump(aita.classifier, 'Aita.model')
fp = open('./assets/ModelParameters.txt', 'w')
fp.write("max_depth=10\nn_estimators=1000\nmin_samples_split=2")
fp.close()
print("------>>> 保存模型结束 <<<------")
def run(path):
# ? 预测数据
aita = Aita()
aita.classifier = joblib.load('Aita.model')
predict_chunk = aita.file_reader(path)
predict_main = aita.get_answers_main(predict_chunk)
for i in range(len(predict_main)):
predict_res = aita.predict(predict_main[i])
print("\n----------------------------")
print("预测结果: " + str(predict_res))
print("----------------------------")
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument("-train", help="训练模型", action="store_true")
parser.add_argument("-predict", help="-train filename.txt")
parser.add_argument("-test", help="测试模型", action="store_true")
args = parser.parse_args()
if args.train:
train()
elif args.predict:
run(args.predict)
elif args.test:
test()
else:
train()
test()
# run()