forked from aisyahbil/NLP_Models
-
Notifications
You must be signed in to change notification settings - Fork 0
/
modelling.py
197 lines (167 loc) · 7.61 KB
/
modelling.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
import numpy as np
import pandas as pd
#from bs4 import BeautifulSoup
import matplotlib.pyplot as plt
#import seaborn as sns
from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer
#from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC
from sklearn.model_selection import train_test_split, StratifiedKFold, cross_val_score
from sklearn.pipeline import make_pipeline, Pipeline
from sklearn.model_selection import GridSearchCV
from sklearn.metrics import make_scorer, accuracy_score, f1_score, roc_curve, auc
from sklearn.metrics import confusion_matrix, roc_auc_score, recall_score, precision_score
from sklearn import preprocessing
#from sklearn.externals import joblib
from sklearn.model_selection import learning_curve
import pickle
from joblib import dump, load
def holdout(data, test_size = 0.2):
train, test = train_test_split(data, test_size=test_size, random_state=1)
X_train = train['cleaned_text'].values
X_test = test['cleaned_text'].values
y_train = train['Label'].values
y_test = test['Label'].values
return X_train, X_test, y_train, y_test
def crossvaldata(data):
X = data['cleaned_text'].values
y = data['Label'].values
return X,y
def vektorisasi(data, termfrequency = True): #data yang sudah di clean
if termfrequency:
tf_vectorizer = CountVectorizer(max_df=1.0, min_df=1)
dtm_tf = tf_vectorizer.fit_transform(data)
tf_terms = tf_vectorizer.get_feature_names()
#print(dtm_tf.shape)
else:
tf_vectorizer = TfidfVectorizer(max_df=1.0, min_df=1)
dtm_tf = tf_vectorizer.fit_transform(data)
tf_terms = tf_vectorizer.get_feature_names()
#print(dtm_tf.shape)
return tf_vectorizer,dtm_tf,tf_terms
def savetokenizer(filepath, tokenizer):
with open(filepath, 'wb') as handle:
pickle.dump(tokenizer, handle, protocol=pickle.HIGHEST_PROTOCOL)
#def savevektorisasi(data, filename):
# with open('vectorizer.pk', 'wb') as fin:
# pickle.dump(vectorizer, fin)
def matrixtermfreq(data, termfrequency = True):
#myvocabulary = tf_terms
corpus = {i : str(data['cleaned_text'].iloc[i]) for i in range(0,len(data))}
if termfrequency:
tf = CountVectorizer(max_df=1, min_df=1)
tfs = tf.fit_transform(corpus.values())
feature_names = tf.get_feature_names()
else:
tfidf = TfidfVectorizer(max_df=1, min_df=1)
tfs = tfidf.fit_transform(corpus.values())
feature_names = tfidf.get_feature_names()
corpus_index = [n for n in corpus]
df = pd.DataFrame(tfs.T.todense(), index=feature_names, columns=corpus_index)
#print(df)
return df
def modelling(data, modelname = str(), crossval = True, termfrequency = False, n_fold = 3, kernel = 'linear', n_jobs=1):
data.cleaned_text = data.cleaned_text.astype('str')
vektor = vektorisasi(data.cleaned_text, termfrequency = termfrequency)
savetokenizer('./NLP_Models/tokenizer_hatespeech/'+ modelname+ '_tokenizer.pickle', vektor[0])
kfolds = StratifiedKFold(n_splits=n_fold, shuffle=True, random_state=1)
np.random.seed(1)
pipeline_svm = make_pipeline(vektor[0],
SVC(probability=True, kernel=kernel, class_weight="balanced"))
grid_svm = GridSearchCV(pipeline_svm,
param_grid = {'svc__C': [0.01, 0.1]},
cv = kfolds,
scoring="roc_auc",
verbose=1,
n_jobs=n_jobs)
if crossval:
X, y = crossvaldata(data)
#grid_svm.fit(X, y)
model = grid_svm.fit(X, y)
score = grid_svm.score(X, y)
print("roc_auc model terbaik adalah:", score)
scorebest = grid_svm.best_estimator_.score(X,y)
print("roc_auc model estimator terbaik adalah:", scorebest)
bestparameter = grid_svm.best_params_
print("Parameter terbaik adalah:", bestparameter)
bestscore = grid_svm.best_score_
print("Rataan roc_auc model tiap fold adalah:" ,bestscore)
filename = './NLP_Models/model_hatespeech/'+ modelname + 'hate_detection.joblib'
dump(model, filename, compress = 1)
else:
X_train, X_test, y_train, y_test = holdout(data)
#grid_svm.fit(X_train, y_train)
model = grid_svm.fit(X_train, y_train)
score = grid_svm.score(X_test, y_test)
print("roc_auc model terbaik adalah:",score)
scorebest = grid_svm.best_estimator_.score(X_test,y_test)
print("roc_auc model estimator terbaik adalah:", scorebest)
bestparameter = grid_svm.best_params_
print("Parameter terbaik adalah:", bestparameter)
bestscore = grid_svm.best_score_
print("Rataan roc_auc model tiap fold adalah:" ,bestscore)
filename = './NLP_Models/model_hatespeech/'+ modelname + 'hate_detection.joblib'
dump(model, filename)
return model, score, bestscore
def confusionMatrix(model, X,y):
y_pred = model.best_estimator_.predict(X)
cm = confusion_matrix(y, y_pred)
return cm
def report_results(model, X, y):
pred_proba = model.predict_proba(X)[:, 1]
pred = model.predict(X)
auC = roc_auc_score(y, pred_proba)
acc = accuracy_score(y, pred)
f1 = f1_score(y, pred, Labels=["-1"], average='micro', pos_Label="-1")
prec = precision_score(y, pred, Labels=["-1"], average='micro', pos_Label="-1")
rec = recall_score(y, pred, Labels=["-1"], average='micro', pos_Label="-1")
return {'auc': auC, 'f1': f1, 'acc': acc, 'precision': prec, 'recall': rec}
def get_roc_curve(model, X, y):
pred_proba = model.predict_proba(X)[:, 1]
fpr, tpr, _ = roc_curve(y, pred_proba)
return fpr, tpr
def visualroc(model,X,y):
roc_svm = get_roc_curve(model, X, y)
fpr, tpr = roc_svm
plt.figure(figsize=(14,8))
plt.plot(fpr, tpr, color="red")
plt.plot([0, 1], [0, 1], color='black', lw=2, linestyle='--')
plt.xlim([0.0, 1.0])
plt.ylim([0.0, 1.05])
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title('Roc curve')
plt.show()
#train_sizes, train_scores, test_scores = \
# learning_curve(model.best_estimator_, X_train, y_train, cv=10, n_jobs=-1,
# scoring="roc_auc", train_sizes=np.linspace(.1, 1.0, 10), random_state=1)
def loadmodel(filename):
model = load(filename)
return model
def plot_learning_curve(X, y, train_sizes, train_scores, test_scores, title='', ylim=None, figsize=(14,8)):
plt.figure(figsize=figsize)
plt.title(title)
if ylim is not None:
plt.ylim(*ylim)
plt.xlabel("Training examples")
plt.ylabel("Score")
train_scores_mean = np.mean(train_scores, axis=1)
train_scores_std = np.std(train_scores, axis=1)
test_scores_mean = np.mean(test_scores, axis=1)
test_scores_std = np.std(test_scores, axis=1)
plt.grid()
plt.fill_between(train_sizes, train_scores_mean - train_scores_std,
train_scores_mean + train_scores_std, alpha=0.1,
color="r")
plt.fill_between(train_sizes, test_scores_mean - test_scores_std,
test_scores_mean + test_scores_std, alpha=0.1, color="g")
plt.plot(train_sizes, train_scores_mean, 'o-', color="r",
label="Training score")
plt.plot(train_sizes, test_scores_mean, 'o-', color="g",
label="Cross-validation score")
plt.legend(loc="lower right")
# return plt
def showlearningcurve( X, y, train_sizes, train_scores, test_scores):
plot_learning_curve(X, y, train_sizes, train_scores, test_scores, ylim=(0.7, 1.01), figsize=(14,6))
plt.show()
# return train_sizes, train_scores, test_scores