-
Notifications
You must be signed in to change notification settings - Fork 0
/
kdataset-preprocessing.py
372 lines (312 loc) · 16.9 KB
/
kdataset-preprocessing.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
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
import os
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import itertools
import json
import random
import argparse
# 데이터 크기 확인 함수
def mem_usage(pandas_obj):
if isinstance(pandas_obj,pd.DataFrame):
usage_b = pandas_obj.memory_usage(deep=True).sum()
else: # we assume if not a df it's a series
usage_b = pandas_obj.memory_usage(deep=True)
usage_mb = usage_b / 1024 ** 2 # convert bytes to megabytes
return "{:03.2f} MB".format(usage_mb)
## 타입별 평균 크기 확인 함수
def type_memory(data) :
for dtype in ['float','int','object']:
selected_dtype = data.select_dtypes(include=[dtype])
mean_usage_b = selected_dtype.memory_usage(deep=True).mean()
mean_usage_mb = mean_usage_b / 1024 ** 2
print("Average memory usage for {} columns: {:03.2f} MB".format(dtype,mean_usage_mb))
## 이산형 데이터 사이즈 축소 함소
def int_memory_reduce(data) :
data_int = data.select_dtypes(include=['int'])
converted_int = data_int.apply(pd.to_numeric,downcast='unsigned')
print(f"Before : {mem_usage(data_int)} -> After : {mem_usage(converted_int)}")
data[converted_int.columns] = converted_int
return data
## 연속형 데이터 사이즈 축소 함소
def float_memory_reduce(data) :
data_float = data.select_dtypes(include=['float'])
converted_float = data_float.apply(pd.to_numeric,downcast='float')
print(f"Before : {mem_usage(data_float)} -> After : {mem_usage(converted_float)}")
data[converted_float.columns] = converted_float
return data
## 문자형 데이터 사이즈 축소 함소
def object_memory_reduce(data) :
gl_obj = data.select_dtypes(include=['object']).copy()
converted_obj = pd.DataFrame()
for col in gl_obj.columns:
num_unique_values = len(gl_obj[col].unique())
num_total_values = len(gl_obj[col])
if num_unique_values / num_total_values < 0.5:
converted_obj.loc[:,col] = gl_obj[col].astype('category')
else:
converted_obj.loc[:,col] = gl_obj[col]
print(f"Before : {mem_usage(gl_obj)} -> After : {mem_usage(converted_obj)}")
data[converted_obj.columns] = converted_obj
return data
def load_dataset(args):
kdf = pd.read_excel(args.dataset_path, sheet_name='rawdata_최종(1129)-250497', engine='openpyxl')
#kdf = kdf.rename(columns=kdf.iloc[0])
#kdf = kdf.drop(kdf.index[0])
print(f"Original Dataset Shape : {kdf.shape}")
return kdf
def generate_userdata(args, kdf):
#### Generating User Dataset
user_data = kdf[["R_id", "r_gender", "age", "mar", "job", "income",
"r_style1", "r_style2", "r_style3", "r_style4", "r_style5"]]
user_data = user_data.drop_duplicates(['R_id']).reset_index(drop=True)
user_data["user"] = user_data.reset_index()["index"]
user_data = user_data[["user", "R_id", "r_gender", "age", "mar", "job", "income",
"r_style1", "r_style2", "r_style3", "r_style4", "r_style5"]]
user_data.columns = ["user", "user_name", "r_gender", "age", "mar", "job", "income",
"r_style1", "r_style2", "r_style3", "r_style4", "r_style5"]
user_data = user_data[["user", "user_name", "r_gender", "age", "mar", "job", "income",
"r_style1", "r_style2", "r_style3", "r_style4", "r_style5"]]
user_data = int_memory_reduce(user_data)
user_data = float_memory_reduce(user_data)
user_data = object_memory_reduce(user_data)
user_data.to_csv(f"{args.save_path}/user_data.csv")
print(f"User Dataset Shape : {user_data.shape}")
return user_data
def generate_itemdata(args, kdf):
#### Generating Item Dataset
# Q2
q2_dict = {"1" : "spring fall", "2" : "summer", "3" : "winter"}
# Q3
q3_dict = {"1" : "attendance",
"2" : "date",
"3" : "event",
"4" : "social gathering",
"5" : "daily",
"6" : "leisure sports",
"7" : "trip vacation",
"8" : "etc"}
# Q411
q411_dict = {"1":"loose","2":"appropriate","3":"tight"}
# Q412
q412_dict = {"1":"dark","2":"bright"}
# Q413
q413_dict = {"1":"cold","2":"warm"}
# Q414
q414_dict = {"1":"heavy","2":"light"}
# Q4201
q4201_dict = {"0":"no","1":"nice"}
# Q4202
q4202_dict = {"0":"no","2":"urban"}
# Q4203
q4203_dict = {"0":"no","3":"trendy"}
# Q4204
q4204_dict = {"0":"no","4":"sophisticated"}
# Q4205
q4205_dict = {"0":"no","5":"clean"}
# Q4206
q4206_dict = {"0":"no","6":"magnificent"}
# Q4207
q4207_dict = {"0":"no","7":"unique"}
# Q4208
q4208_dict = {"0":"no","8":"easy"}
# Q4209
q4209_dict = {"0":"no","9":"open mined"}
# Q4210
q4210_dict = {"0":"no","10":"practical"}
# Q4211
q4211_dict = {"0":"no","11":"activity"}
# Q4212
q4212_dict = {"0":"no","12":"comfortable"}
# Q4213
q4213_dict = {"0":"no","13":"bubbly"}
# Q4214
q4214_dict = {"0":"no","14":"feminine"}
# Q4215
q4215_dict = {"0":"no","15":"manly"}
# Q4216
q4216_dict = {"0":"no","16":"soft"}
item_data = kdf[['imgName','era','style','gender',
'Q1','Q2','Q3','Q411','Q412','Q413','Q414','Q4201','Q4202','Q4203','Q4204','Q4205',
'Q4206','Q4207','Q4208','Q4209','Q4210','Q4211','Q4212','Q4213','Q4214','Q4215','Q4216','Q5']]
item_data['Q414'] = item_data['Q414'].fillna(item_data['Q414'].mode()[0])
for col_i in item_data.columns.tolist()[4:]:
item_data.loc[:, col_i] = item_data.loc[:, col_i].astype("int")
for col_i in ['Q2','Q3','Q411','Q412','Q413','Q414','Q4201','Q4202','Q4203','Q4204','Q4205',
'Q4206','Q4207','Q4208','Q4209','Q4210','Q4211','Q4212','Q4213','Q4214','Q4215','Q4216']:
item_data.loc[:, col_i] = item_data.loc[:, col_i].astype("str")
item_data["season"] = [q2_dict[str(i)] for i in item_data["Q2"].values.tolist()]
item_data["tpo"] = [q3_dict[str(i)] for i in item_data["Q3"].values.tolist()]
item_data["fit"] = [q411_dict[str(i)] for i in item_data["Q411"].values.tolist()]
item_data["brightness"] = [q412_dict[str(i)] for i in item_data["Q412"].values.tolist()]
item_data["temperature"] = [q413_dict[str(i)] for i in item_data["Q413"].values.tolist()]
item_data["weight"] = [q414_dict[str(i)] for i in item_data["Q414"].values.tolist()]
item_data["nice"] = [q4201_dict[str(i)] for i in item_data["Q4201"].values.tolist()]
item_data["urban"] = [q4202_dict[str(i)] for i in item_data["Q4202"].values.tolist()]
item_data["trendy"] = [q4203_dict[str(i)] for i in item_data["Q4203"].values.tolist()]
item_data["sophisticated"] = [q4204_dict[str(i)] for i in item_data["Q4204"].values.tolist()]
item_data["clean"] = [q4205_dict[str(i)] for i in item_data["Q4205"].values.tolist()]
item_data["magnificent"] = [q4206_dict[str(i)] for i in item_data["Q4206"].values.tolist()]
item_data["unique"] = [q4207_dict[str(i)] for i in item_data["Q4207"].values.tolist()]
item_data["easy"] = [q4208_dict[str(i)] for i in item_data["Q4208"].values.tolist()]
item_data["open"] = [q4209_dict[str(i)] for i in item_data["Q4209"].values.tolist()]
item_data["practical"] = [q4210_dict[str(i)] for i in item_data["Q4210"].values.tolist()]
item_data["activity"] = [q4211_dict[str(i)] for i in item_data["Q4211"].values.tolist()]
item_data["comfortable"] = [q4212_dict[str(i)] for i in item_data["Q4212"].values.tolist()]
item_data["bubbly"] = [q4213_dict[str(i)] for i in item_data["Q4213"].values.tolist()]
item_data["feminine"] = [q4214_dict[str(i)] for i in item_data["Q4214"].values.tolist()]
item_data["manly"] = [q4215_dict[str(i)] for i in item_data["Q4215"].values.tolist()]
item_data["soft"] = [q4216_dict[str(i)] for i in item_data["Q4216"].values.tolist()]
item_data = item_data[["imgName", "era", "style", "gender", 'season', 'tpo', 'fit', 'brightness',
'temperature', 'weight', 'nice', 'urban', 'trendy', 'sophisticated',
'clean', 'magnificent', 'unique', 'easy', 'open', 'practical',
'activity', 'comfortable', 'bubbly', 'feminine', 'manly', 'soft', "Q1", "Q5"]]
item_data = pd.get_dummies(item_data, columns = ['season', "tpo", 'fit', 'brightness',
'temperature', 'weight', 'nice', 'urban', 'trendy', 'sophisticated',
'clean', 'magnificent', 'unique', 'easy', 'open', 'practical',
'activity', 'comfortable', 'bubbly', 'feminine', 'manly', 'soft'])
item_data = item_data.groupby(["imgName", "era", "style", "gender"]).sum().reset_index()
item_data.columns = ["item_name", "era", "style", "gender"] + item_data.columns.tolist()[4:]
item_data = int_memory_reduce(item_data)
item_data = float_memory_reduce(item_data)
item_data = object_memory_reduce(item_data)
# season
spring_and_fall = item_data[item_data["season_spring fall"]>=1].index
summer = item_data[item_data["season_summer"]>=1].index
winter = item_data[item_data["season_winter"]>=1].index
# tpo
attend = item_data[item_data["tpo_attendance"]>=1].index
daily = item_data[item_data["tpo_daily"]>=1].index
date = item_data[item_data["tpo_date"]>=1].index
etc = item_data[item_data["tpo_etc"]>=1].index
event = item_data[item_data["tpo_event"]>=1].index
sports = item_data[item_data["tpo_leisure sports"]>=1].index
social = item_data[item_data["tpo_social gathering"]>=1].index
trip = item_data[item_data["tpo_trip vacation"]>=1].index
# fit
appro = item_data[item_data["fit_appropriate"]>=1].index
loo = item_data[item_data["fit_loose"]>=1].index
tig = item_data[item_data["fit_tight"]>=1].index
# brightness
bright = item_data[item_data["brightness_bright"]>=1].index
dark = item_data[item_data["brightness_dark"]>=1].index
# temperature
cold = item_data[item_data["temperature_cold"]>=1].index
warm = item_data[item_data["temperature_warm"]>=1].index
# weight
heavy = item_data[item_data["weight_heavy"]>=1].index
light = item_data[item_data["weight_light"]>=1].index
nice_no = item_data[item_data["nice_no"]>=1].index
nice_nice = item_data[item_data["nice_nice"]>=1].index
urban_no = item_data[item_data["urban_no"]>=1].index
urban_urban = item_data[item_data["urban_urban"]>=1].index
trendy_no = item_data[item_data["trendy_no"]>=1].index
trendy_trendy = item_data[item_data["trendy_trendy"]>=1].index
sophisticated_no = item_data[item_data["sophisticated_no"]>=1].index
sophisticated_sohp = item_data[item_data["sophisticated_sophisticated"]>=1].index
clean_no = item_data[item_data["clean_no"]>=1].index
clean_clean = item_data[item_data["clean_clean"]>=1].index
magnificent_no = item_data[item_data["magnificent_no"]>=1].index
magnificent_magnificent = item_data[item_data["magnificent_magnificent"]>=1].index
unique_no = item_data[item_data["unique_no"]>=1].index
unique_unique = item_data[item_data["unique_unique"]>=1].index
easy_no = item_data[item_data["easy_no"]>=1].index
easy_easy = item_data[item_data["easy_easy"]>=1].index
open_no = item_data[item_data["open_no"]>=1].index
open_open = item_data[item_data["open_open mined"]>=1].index
practical_no = item_data[item_data["practical_no"]>=1].index
practical_practical = item_data[item_data["practical_practical"]>=1].index
activity_no = item_data[item_data["activity_no"]>=1].index
activity_activity = item_data[item_data["activity_activity"]>=1].index
comfortable_no = item_data[item_data["comfortable_no"]>=1].index
comfortable_comfortable = item_data[item_data["comfortable_comfortable"]>=1].index
bubbly_no = item_data[item_data["bubbly_no"]>=1].index
bubbly_bubbly = item_data[item_data["bubbly_bubbly"]>=1].index
feminine_no = item_data[item_data["feminine_no"]>=1].index
feminine_feminine = item_data[item_data["feminine_feminine"]>=1].index
manly_no = item_data[item_data["manly_no"]>=1].index
manly_manly = item_data[item_data["manly_manly"]>=1].index
soft_no = item_data[item_data["soft_no"]>=1].index
soft_soft = item_data[item_data["soft_soft"]>=1].index
item_data["season"] = ""
item_data["tpo"] = ""
item_data["fit"] = ""
item_data["brightness"] = ""
item_data["temperature"] = ""
item_data["weight"] = ""
item_data.iloc[spring_and_fall, -6] = "spring fall"
item_data.iloc[summer, -6] = item_data.iloc[summer, -6] + " summer"
item_data.iloc[winter, -6] = item_data.iloc[winter, -6] + " winter"
item_data.iloc[attend, -5] = "attendance"
item_data.iloc[daily, -5] = item_data.iloc[daily, -5] + " daily"
item_data.iloc[date, -5] = item_data.iloc[date, -5] + " date"
item_data.iloc[etc, -5] = item_data.iloc[etc, -5] + " etc"
item_data.iloc[event, -5] = item_data.iloc[event, -5] + " event"
item_data.iloc[sports, -5] = item_data.iloc[sports, -5] + " leisure sports"
item_data.iloc[social, -5] = item_data.iloc[social, -5] + " social gathering"
item_data.iloc[trip, -5] = item_data.iloc[trip, -5] + " social trip vacation"
item_data.iloc[appro, -4] = "appropriate"
item_data.iloc[loo, -4] = item_data.iloc[loo, -4] + " loose"
item_data.iloc[tig, -4] = item_data.iloc[tig, -4] + " tight"
item_data.iloc[bright, -3] = item_data.iloc[bright, -3] + " bright"
item_data.iloc[dark, -3] = item_data.iloc[dark, -3] + " dark"
item_data.iloc[cold, -2] = item_data.iloc[cold, -2] + " cold"
item_data.iloc[warm, -2] = item_data.iloc[warm, -2] + " warm"
item_data.iloc[heavy, -1] = item_data.iloc[heavy, -1] + " heavy"
item_data.iloc[light, -1] = item_data.iloc[light, -1] + " light"
item_data["item"] = item_data.reset_index()["index"]
item_data = \
item_data[['item','item_name', 'era', 'style', 'gender',
'season', 'tpo', 'fit', 'brightness', 'temperature', 'weight',
'nice_nice', 'nice_no', 'urban_no', 'urban_urban', 'trendy_no',
'trendy_trendy', 'sophisticated_no', 'sophisticated_sophisticated',
'clean_clean', 'clean_no', 'magnificent_magnificent', 'magnificent_no',
'unique_no', 'unique_unique', 'easy_easy', 'easy_no', 'open_no',
'open_open mined', 'practical_no', 'practical_practical',
'activity_activity', 'activity_no', 'comfortable_comfortable',
'comfortable_no', 'bubbly_bubbly', 'bubbly_no', 'feminine_feminine',
'feminine_no', 'manly_manly', 'manly_no', 'soft_no', 'soft_soft']]
item_data['season'] = item_data['season'].str.strip()
item_data['tpo'] = item_data['tpo'].str.strip()
item_data['fit'] = item_data['fit'].str.strip()
item_data['brightness'] = item_data['brightness'].str.strip()
item_data['temperature'] = item_data['temperature'].str.strip()
item_data['weight'] = item_data['weight'].str.strip()
for re_col in item_data.columns.tolist()[11:]:
item_data.loc[item_data[re_col] >= 1, re_col] = 1
item_data = int_memory_reduce(item_data)
item_data = float_memory_reduce(item_data)
item_data = object_memory_reduce(item_data)
item_data.to_csv(f"{args.save_path}/item_data.csv")
print(f"Item Dataset Shape : {item_data.shape}")
return item_data
def generate_ratedata(args, kdf, user_data, item_data):
rate_data = kdf[["R_id", "imgName", "Q1", "Q5"]]
rate_data.columns = ["user", "item", "Q1", "Q5"]
u_dict = user_data[["user", "user_name"]].astype("str").set_index("user_name")
u_dict = u_dict.T.to_dict('records')[0]
i_dict = item_data[["item", "item_name"]].astype("str").set_index("item_name")
i_dict = i_dict.T.to_dict('records')[0]
#rate_data["user"] = [int(u_dict[str(i)]) for i in rate_data["user"].values.tolist()]
rate_data["item"] = [int(i_dict[str(i)]) for i in rate_data["item"].values.tolist()]
rate_data_q1q2 = rate_data[["user", "item", "Q1", "Q5"]]
rate_data_q1q2["rate"] = rate_data_q1q2[["Q1", "Q5"]].mean(axis=1)
rate_data_q1q2 = rate_data_q1q2[["user", "item", "rate"]]
rate_data_q1q2 = int_memory_reduce(rate_data_q1q2)
rate_data_q1q2 = float_memory_reduce(rate_data_q1q2)
rate_data_q1q2.to_csv(f"{args.save_path}/rate_data.csv")
print(f"Rating Dataset Shape : {rate_data_q1q2.shape}")
def build_data(args):
kdf = load_dataset(args)
user_data = generate_userdata(args, kdf)
item_data = generate_itemdata(args, kdf)
generate_ratedata(args, kdf, user_data, item_data)
print('\033[95m' + f'----complete----' + '\033[0m')
if __name__ == '__main__':
# Arguments
parser = argparse.ArgumentParser()
parser.add_argument('--dataset-path', type=str, default="KData/패션 선호도조사 결과_최종data(1129).xlsx")
parser.add_argument('--save-path', type=str, default='KData')
args = parser.parse_args()
# Training
build_data(args)