-
Notifications
You must be signed in to change notification settings - Fork 0
/
pre_analyze.py
220 lines (192 loc) · 6.56 KB
/
pre_analyze.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
219
220
import data
import csv
import math
import numpy as np
import visualization
import scipy
from sklearn.cross_validation import train_test_split
import random
def csv_pre_train(filename):
f = file(filename, 'rU')
reader = csv.reader(f)
headers = reader.next()
data = []
for index, row in enumerate(reader):
temp = []
for item in row:
item = item.strip()
temp.append(item)
data.append(temp[:])
len_pre_train = int(len(data)*1/4)
s1 = filename[:-4] + "Pre.csv"
f1 = open(s1, 'ab')
csvWriter1 = csv.writer(f1)
csvWriter1.writerow(headers)
for i in range(len(data[:len_pre_train])):
csvWriter1.writerow(data[i])
return s1
def csv_data_key(filename, savefile):
f1 = file(filename, 'rU')
f2 = open(savefile, 'ab')
csvwriter = csv.writer(f2)
reader = csv.reader(f1)
headers = reader.next()[:-1]
typed = []
for i in range(len(headers)):
typed.append("numeric")
d = []
for index, row in enumerate(reader):
temp = []
for item in row[:-1]:
if item == " ":
item = np.nan
item = item.strip()
temp.append(item)
d.append(temp[:])
csvwriter.writerow(headers)
csvwriter.writerow(typed)
for i in range(len(d)):
csvwriter.writerow(d[i])
def create_cat(data, column_name):
col = data.get_data([column_name])
pixel_li = data.get_image_data()
def determine_eye_size(filename1):
d1 = data.Data(filename1)
cx = d1.get_data(["left_eye_center_x"])
icx = d1.get_data(["left_eye_inner_corner_x"])
ocx = d1.get_data(["left_eye_outer_corner_x"])
cy = d1.get_data(["left_eye_center_y"])
icy = d1.get_data(["left_eye_inner_corner_y"])
ocy = d1.get_data(["left_eye_outer_corner_y"])
range_x = ocx - icx
range_y = ocy -icy
print "left_eye_inner_corner_x_mean", np.nanmean(icx)
print "left_eye_outer_corner_x_mean", np.nanmean(ocx)
print "left_eye_center_x_mean", np.nanmean(cx)
print "range_x mean", np.nanmean(range_x)
print "range_x std dev", np.nanstd(range_x)
print "left_eye_inner_corner_y_mean", np.nanmean(icy)
print "left_eye_outer_corner_y_mean", np.nanmean(ocy)
print "left_eye_center_y_mean", np.nanmean(cy)
print "range_y mean", np.nanmean(range_y)
print "range_y std dev", np.nanstd(range_y)
def create_data_clustering(data, filenameTrain, filenameCat):
f = open(filenameTrain, 'ab')
f2 = open(filenameCat, 'ab')
csvWriter1 = csv.writer(f)
csvWriter2 = csv.writer(f2)
header = []
category_dic = {}
typed = []
for row in range(96):
for columns in range(96):
if row+1 < 10:
sr = str(0) + str(row+1)
else:
sr = str(row+1)
if columns + 1<10:
sc = str(0) + str(columns+1)
else:
sc = str(columns+1)
s1 = "pixel_" + sr + sc
header.append(s1)
typed.append("numeric")
csvWriter1.writerow(header)
csvWriter1.writerow(typed)
header2 = ["left_eye_center_x", "left_eye_center_y"]
typed = ["numeric", "numeric"]
#I divide image into 36 categories each category has 16*16 pixel
cat = data.get_data(["left_eye_center_x", "left_eye_center_y"]).tolist()
csvWriter2.writerow(header2)
csvWriter2.writerow(typed)
image = data.get_image_data()
#print "image", image
for i in range(len(image)):
csvWriter1.writerow(image[i])
csvWriter2.writerow(cat[i])
#print "data.get_data(["Image"])",data.get_data(["Image"])
f.close()
f2.close()
def cross_validation_m(fileData, fileCat, train_ratio, test_ratio):
dfile = data.Data(fileData)
catfile = data.Data(fileCat)
dfile_data = dfile.get_data(dfile.get_headers())
catfile_data = catfile.get_data(catfile.get_headers())
X_train, X_test, y_train, y_test = train_test_split( dfile_data, catfile_data, test_size=test_ratio, random_state=0)
train_num = int(dfile_data.shape[0] * train_ratio)
train_num_indecies = random.sample(range(X_train.shape[0]), train_num)
X_train = X_train[train_num_indecies, :]
y_train = y_train[train_num_indecies, :]
return X_train, X_test, y_train, y_test
def cross_validation(fileData, fileCat, train_ratio, test_ratio, s1, s2, s3, s4):
dfile = data.Data(fileData)
catfile = data.Data(fileCat)
dfile_data = dfile.get_data(dfile.get_headers())
catfile_data = catfile.get_data(catfile.get_headers())
ftrainX = open(s1, 'ab')
ftrainY = open(s2, 'ab')
ftestX = open(s3, 'ab')
ftestY = open(s4, 'ab')
trainX = csv.writer(ftrainX)
trainY = csv.writer(ftrainY)
testX = csv.writer(ftestX)
testY = csv.writer(ftestY)
#writing header to trainX and testX
trainX.writerow(dfile.get_headers())
testX.writerow(dfile.get_headers())
#writing types to trainX and testX
trainX.writerow(dfile.get_raw_types())
testX.writerow(dfile.get_raw_types())
#writing header to trainY and testY
trainY.writerow(catfile.get_headers())
testY.writerow(catfile.get_headers())
#writing type to trainY and testY
trainY.writerow(catfile.get_raw_types())
testY.writerow(catfile.get_raw_types())
X_train, X_test, y_train, y_test = train_test_split( dfile_data, catfile_data, test_size=test_ratio, random_state=0)
#create testX and testY
testX_list = X_test.tolist()
testY_list = y_test.tolist()
for i in range(len(testY_list)):
testX.writerow(testX_list[i])
testY.writerow(testY_list[i])
#create trainX and tainY
train_num = int(dfile_data.shape[0] * train_ratio)
train_num_indecies = random.sample(range(X_train.shape[0]), train_num)
trainX_list = X_train.tolist()
trainY_list = y_train.tolist()
for num in train_num_indecies:
trainX.writerow(trainX_list[num])
trainY.writerow(trainY_list[num])
ftrainX.close()
ftrainY.close()
ftestX.close()
ftestY.close()
def pre_analyze_preparation():
trainingdata = data.FacialData("training.csv")
create_data_clustering(trainingdata, "dataX.csv", "dataY.csv")
def pre_analyze_param(trainratio, testratio):
X_train, X_test, y_train, y_test = cross_validation_m("dataX.csv", "dataY.csv", trainratio, testratio)
return X_train, X_test, y_train, y_test
def main():
#f = csv_pre_train("training.csv")
#pixelInfo = visualization.show_pictures_data("training.csv")
#visualization.show_picture(pixelInfo)
#traingdata = data.FacialData("trainingPre.csv")
#create_data_clustering(traingdata, "trainingPreX.csv", "trainingPreY.csv")
#trainingdata = data.FacialData("training.csv")
#create_data_clustering(trainingdata, "dataX.csv", "dataY.csv")
d = data.Data("traindataX.csv")
#pixel = d.get_data(d.get_headers())
#print "pixel.shape", pixel.shape
#pixel = pixel.tolist()
#print "len(pixelInfo)", len(pixelInfo)
#print "len(pixel)", len(pixel.tolist())
#print "pixelInfo[0]", pixelInfo[0]
#print "pixel.tolist()[0]",pixel.tolist()[0]
#visualization.show_picture(pixel)
normalize_by_example(d, filename="trainNorm.csv")
#csv_data_key("training.csv", "trainingKey.csv")
#determine_eye_size("trainingKey.csv")
if __name__ == "__main__":
main()