DRecSys/pre-processing.py at master · Talessil/DRecSys · GitHub

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
import pandas as pd
import numpy as np
import csv
from sklearn.utils import shuffle

"""
@author: Talessil
preprocessing code: outliers scale reduction, normalization,
                        instance splitting - test and training (optional)
input: fp_input.csv
output: fp_input_def.csv
"""

# READING
data = pd.read_csv("fp_input.csv", sep=";", header=0)
array_all = data.values
array = data.values

size = 0
for n in array:
    size = size + 1

""" OUTLIERS SCALE REDUCTION : FOR EACH ATTRIBUTE """

# 3rd QUARTILE CALCULUS

q1, q3= np.percentile(data.discussion,[25,75])
iqr = q3 - q1
lower_bound_discussion = q1 -(1.5 * iqr)
upper_bound_discussion = q3 +(1.5 * iqr)
print(lower_bound_discussion)
print(upper_bound_discussion)

q1, q3= np.percentile(data.review,[25,75])
iqr = q3 - q1
lower_bound_review = q1 -(1.5 * iqr)
upper_bound_review = q3 +(1.5 * iqr)
print(lower_bound_review)
print(upper_bound_review)

q1, q3= np.percentile(data.qntags,[25,75])
iqr = q3 - q1
lower_bound_qntags = q1 -(1.5 * iqr)
upper_bound_qntags = q3 +(1.5 * iqr)
print(lower_bound_qntags)
print(upper_bound_qntags)

q1, q3= np.percentile(data.pull,[25,75])
iqr = q3 - q1
lower_bound_pull = q1 -(1.5 * iqr)
upper_bound_pull = q3 +(1.5 * iqr)
print(lower_bound_pull)
print(upper_bound_pull)

# RE-WRITTING WITH THE NEW CHANGES (TEMP FILE IS CREATED FOR TEMPORARY CHANGES)

# REMOVE TARGET ATTRIBUTE
k = 0
with open('temp.csv', mode='w') as reduct:
    reduct = csv.writer(reduct, delimiter=';', quotechar='"', quoting=csv.QUOTE_NONE,lineterminator = '\n')
    reduct.writerow(['discussion', 'review', 'qntags', 'pull'])
    for k in range(size):
        reduct.writerow([str(array[k][1]),str(array[k][2]),str(array[k][3]),str(array[k][4])])

# ATTRIBUTE DISCUSSION
data = pd.read_csv("temp.csv", sep=";", header=0)
array = data.values
k = 0
with open('temp.csv', mode='w') as reduct:
    reduct = csv.writer(reduct, delimiter=';', quotechar='"', quoting=csv.QUOTE_NONE,lineterminator = '\n')
    reduct.writerow(['discussion', 'review', 'qntags', 'pull'])
    for k in range(size):
        if(array[k][0]>upper_bound_discussion):
            coef = upper_bound_discussion/array[k][0]
            reduct.writerow([str(upper_bound_discussion),str("%.2f" % (coef*array[k][1])),str("%.2f" % (coef*array[k][2])),str("%.2f" % (coef*array[k][3]))])
        else:
            reduct.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3])])

# ATTRIBUTE REVIEW
data = pd.read_csv("temp.csv", sep=";", header=0)
array = data.values
k = 0
with open('temp.csv', mode='w') as reduct:
    reduct = csv.writer(reduct, delimiter=';', quotechar='"', quoting=csv.QUOTE_NONE,lineterminator = '\n')
    reduct.writerow(['discussion', 'review', 'qntags', 'pull'])
    for k in range(size):
        if(array[k][1]>upper_bound_review):
            coef = upper_bound_review/array[k][1]
            reduct.writerow([str("%.2f" % (coef*array[k][0])),str(upper_bound_review),str("%.2f" % (coef*array[k][2])),str("%.2f" % (coef*array[k][3]))])
        else:
            reduct.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3])])

# ATTRIBUTE QNTAGS
data = pd.read_csv("temp.csv", sep=";", header=0)
array = data.values
k = 0
with open('temp.csv', mode='w') as reduct:
    reduct = csv.writer(reduct, delimiter=';', quotechar='"', quoting=csv.QUOTE_NONE,lineterminator = '\n')
    reduct.writerow(['discussion', 'review', 'qntags', 'pull'])
    for k in range(size):
        if(array[k][2]>upper_bound_qntags):
            coef = upper_bound_qntags/array[k][2]
            reduct.writerow([str("%.2f" % (coef*array[k][0])),str("%.2f" % (coef*array[k][1])),str(upper_bound_qntags),str("%.2f" % (coef*array[k][3]))])
        else:
            reduct.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3])])

# ATTRIBUTE PULLREQUEST
data = pd.read_csv("temp.csv", sep=";", header=0)
array = data.values
k = 0
with open('temp.csv', mode='w') as reduct:
    reduct = csv.writer(reduct, delimiter=';', quotechar='"', quoting=csv.QUOTE_NONE,lineterminator = '\n')
    reduct.writerow(['discussion', 'review', 'qntags', 'pull'])
    for k in range(size):
        if(array[k][3]>upper_bound_pull):
            coef = upper_bound_pull/array[k][3]
            reduct.writerow([str("%.2f" % (coef*array[k][0])),str("%.2f" % (coef*array[k][1])),str("%.2f" % (coef*array[k][2])),str(upper_bound_pull)])
        else:
            reduct.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3])])

############################################################################################################################################################

""" NORMALIZATION """

aux1 = 0
aux2 = 0
aux3 = 0
aux4 = 0
data = pd.read_csv("temp.csv", sep=";", header=0)
array = data.values
for k in range(size):
    if(aux1 < array[k][0]):
        aux1 = array[k][0]
    if(aux2 < array[k][1]):
        aux2 = array[k][1]
    if(aux3 < array[k][2]):
        aux3 = array[k][2]
    if(aux4 < array[k][3]):
        aux4 = array[k][3]

# NORMALIZE
with open('temp.csv', mode='w') as reduct:
    reduct = csv.writer(reduct, delimiter=';', quotechar='"', quoting=csv.QUOTE_NONE,lineterminator = '\n')
    reduct.writerow(['author_id', 'discussion', 'review', 'qntags', 'pull', 'requested'])
    for k in range(size):
        reduct.writerow([str(array_all[k][0]),str("%.2f" % (array[k][0]/aux1)),str("%.2f" % (array[k][1]/aux2)),str("%.2f" % (array[k][2]/aux3)),str("%.2f" % (array[k][3]/aux4)),str(array_all[k][5])])

############################################################################################################################################################

""" SPLIT INSTANCE : TEST AND TRAINING - OPTIONAL"""
"""
cont_req = 0
cont_n_req = 0
data = pd.read_csv("temp.csv", sep=";", header=0)
array = data.values
array = shuffle(array) #SHUFFLE INSTANCE
for k in range(size):
    if array[k][5]==1:
        cont_req = cont_req + 1
    if array[k][5]==0:
        cont_n_req = cont_n_req + 1

coef_test_req = 0.3 * cont_req
coef_test_n_req = 0.3 * cont_n_req

cont_req = 0
cont_n_req = 0
with open('test.csv', mode='w') as test:
    with open('training.csv', mode='w') as training:
        test = csv.writer(test, delimiter=';', quotechar='"', quoting=csv.QUOTE_NONE,lineterminator = '\n')
        test.writerow(['author_id', 'discussion', 'review', 'qntags', 'pull', 'requested'])
        training = csv.writer(training, delimiter=';', quotechar='"', quoting=csv.QUOTE_NONE,lineterminator = '\n')
        training.writerow(['author_id', 'discussion', 'review', 'qntags', 'pull', 'requested'])
        for k in range(size):
            if array[k][5]==1 and cont_req<coef_test_req:
                test.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3]),str(array[k][4]),str(0)])
                cont_req = cont_req + 1
            elif array[k][5]==0 and cont_n_req<coef_test_n_req:
                test.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3]),str(array[k][4]),str(1)])
                cont_n_req = cont_n_req + 1
            else:
                if array[k][5]==1:
                    training.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3]),str(array[k][4]),str(0)])
                else:
                    training.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3]),str(array[k][4]),str(1)])
"""
############################################################################################################################################################

""" WRITE DEFINITIVE FILE"""

data = pd.read_csv("temp.csv", sep=";", header=0)
array = data.values
array = shuffle(array) #SHUFFLE
cont = 0
for k in range(size):
    if array[k][5]==1:
        cont = cont + 1

with open('fp_input_def.csv', mode='w') as defi:
    defi = csv.writer(defi, delimiter=';', quotechar='"', quoting=csv.QUOTE_NONE,lineterminator = '\n')
    defi.writerow(['author_id', 'discussion', 'review', 'qntags', 'pull', 'requested'])
    for k in range(size):
        if array[k][5]==1:
            defi.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3]),str(array[k][4]),str(0)])
        else:
            defi.writerow([str(array[k][0]),str(array[k][1]),str(array[k][2]),str(array[k][3]),str(array[k][4]),str(1)])