-
Notifications
You must be signed in to change notification settings - Fork 0
/
EngineTest.py
198 lines (164 loc) · 5.67 KB
/
EngineTest.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
import pandas as pd
import re
import sys
import math
import numpy as np
from numpy import array
from nltk.corpus import stopwords
from sklearn.feature_extraction.text import CountVectorizer
import pickle
import mysql.connector
from mysql.connector import errorcode
import nltk.stem
english_stemmer = nltk.stem.SnowballStemmer('english')
np.set_printoptions(threshold=sys.maxsize)
try:
cnx = mysql.connector.connect(user='root', password='',
host='127.0.0.1',
database='ta-quran-0-4800')
cursor = cnx.cursor(buffered=True)
except mysql.connector.Error as err:
if err.errno == errorcode.ER_ACCESS_DENIED_ERROR:
print("Something is wrong with your user name or password")
elif err.errno == errorcode.ER_BAD_DB_ERROR:
print("Database does not exist")
else:
print(err)
class StemmedCountVectorizer(CountVectorizer):
def build_analyzer(self):
analyzer = super(StemmedCountVectorizer, self).build_analyzer()
return lambda doc: ([english_stemmer.stem(w) for w in analyzer(doc)])
#fungsi getFeatures
def getFeatures(data):
vectorizer = StemmedCountVectorizer(
# min_df = 1,
analyzer = "word",
tokenizer = None,
preprocessor = None,
stop_words = None,
max_features = 15000
)
dataFeatures = vectorizer.fit_transform(data)
dataFeatures = dataFeatures.toarray()
vocab = vectorizer.get_feature_names()
dist = np.sum(dataFeatures, axis=0)
wordArray = []
dataCount = []
for tag, count in zip(vocab,dist):
wordArray.append(tag)
dataCount.append(count)
return dataCount, wordArray
# get semua data class
def getAllClassList():
className = []
query = ("SELECT DISTINCT level_1 from ta_kelas")
cursor.execute(query,)
temp = []
for level_1, in cursor:
temp.append(level_1)
return temp
# untuk get berapa mutual yang akan digunakan
def getMutual(thresh):
query = ("SELECT parent, child, prob FROM ta_mutual ORDER BY prob DESC")
cursor.execute(query)
p = []
c = []
for parent, child, prob in cursor:
if (prob >= thresh):
if ((parent not in p) and (child not in c)) and ((parent not in c) and (child not in p)):
p.append(parent)
c.append(child)
return p, c
def preProcessing(terjemahan):
letters_only = re.sub("[^a-zA-Z]"," ",terjemahan)
words = letters_only.lower().split()
stops = set(stopwords.words("english"))
real_words = [w for w in words if not w in stops]
return(" ".join(real_words))
# ambil data test
def makeDataSet(rangeawal, rangeakhir):
query = ("SELECT id, teksayat FROM ta_ayat WHERE id > %s AND id <= %s")
cursor.execute(query,(rangeawal,rangeakhir))
id_data_set = []
data_training = []
rows = cursor.fetchall()
for row in rows:
id_data_set.append(row[0])
data_training.append(row[1])
sz_dtTraining = len(data_training)
clear_data = []
for i in range(0,sz_dtTraining):
clear_data.append(preProcessing(data_training[i]))
return id_data_set, clear_data
# misal input ayat "beneficent merciful"
def posterior(arrCh, id_data_set, data_test, level_1):
for i in range(0, len(data_test)):
if data_test[i] != '':
query = ("SELECT SUM(TF), SUM(FF) FROM ta_likelihood WHERE level_1 = %s")
cursor.execute(query,(level_1,))
prob = cursor.fetchone()
sumAllTrue = prob[0]
sumAllFalse = prob[1]
count, splitTrain = getFeatures([data_test[i]])
for j in range(0, len(splitTrain)):
query = ("SELECT TT, TF, FT, FF FROM ta_likelihood WHERE word = %s AND level_1 = %s")
cursor.execute(query,(splitTrain[j], level_1))
cpt = cursor.fetchone()
TTWord = cpt[0] * count[j]
TFWord = cpt[1] * count[j]
FTWord = cpt[2] * count[j]
FFWord = cpt[3] * count[j]
if splitTrain[j] in arrCh:
sumAllTrue = (sumAllTrue - TFWord)
sumAllFalse = (sumAllFalse - FFWord)
elif splitTrain[j] not in arrCh:
sumAllTrue = (sumAllTrue - TFWord) + TTWord
sumAllFalse = (sumAllFalse - FFWord) + FTWord
for k in range(0, len(arrCh)):
query = ("SELECT parent, child, TTT, TTF, TFT, TFF, FTT, FTF, FFT, FFF FROM ta_cpt_split WHERE child = %s and level_1 = %s")
cursor.execute(query,(arrCh[k], level_1))
prob = cursor.fetchone()
TTT = prob[2] * count[j]
TTF = prob[3] * count[j]
TFT = prob[4] * count[j]
TFF = prob[5] * count[j]
FTT = prob[6] * count[j]
FTF = prob[7] * count[j]
FFT = prob[8] * count[j]
FFF = prob[9] * count[j]
if (prob[0] in splitTrain) and (prob[1] in splitTrain):
sumAllTrue = sumAllTrue + TTT
sumAllFalse = sumAllFalse + FTT
elif (prob[0] in splitTrain) and (prob[1] not in splitTrain):
sumAllTrue = sumAllTrue + TTF
sumAllFalse = sumAllFalse + FTF
elif (prob[0] not in splitTrain) and (prob[1] in splitTrain):
sumAllTrue = sumAllTrue + TFT
sumAllFalse = sumAllFalse + FFT
elif (prob[0] not in splitTrain) and (prob[1] not in splitTrain):
sumAllTrue = sumAllTrue + TFF
sumAllFalse = sumAllFalse + FFF
query = ("SELECT prior_yes, prior_no FROM ta_prior WHERE level_1 = %s")
cursor.execute(query,(level_1,))
prior = cursor.fetchone()
posterior_yes = sumAllTrue + prior[0]
posterior_no = sumAllFalse + prior[1]
if posterior_yes > posterior_no:
query = ("INSERT INTO ta_newoutput_split VALUES (%s, %s)")
try:
cursor.execute(query,(id_data_set[i], level_1))
cnx.commit()
except:
cnx.rollback()
# MAIN
rangeawal = 4800
rangeakhir = 6236
arrKelas = getAllClassList()
id_data_set, data_test = makeDataSet(rangeawal, rangeakhir)
arrCh, arrP = getMutual(4)
j = 0
for i in range(0, len(arrKelas)):
j = j+1
print (arrKelas[i])
level_1 = arrKelas[i]
posterior(arrCh, id_data_set, data_test, level_1)