-
Notifications
You must be signed in to change notification settings - Fork 0
/
cleaning.py
268 lines (172 loc) · 12.5 KB
/
cleaning.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
def fill_bin_num(dataframe, feature, bin_feature, bin_size, stat_measure, min_bin=None, max_bin=None, default_val='No'):
if min_bin is None:
min_bin = dataframe[bin_feature].min()
if max_bin is None:
max_bin = dataframe[bin_feature].max()
new_dataframe = dataframe.copy()
df_meancat = pd.DataFrame(columns=['interval', 'stat_measure'])
for num_bin, subset in dataframe.groupby(pd.cut(dataframe[bin_feature], np.arange(min_bin, max_bin+bin_size, bin_size), include_lowest=True)):
if stat_measure is 'mean':
row = [num_bin, subset[feature].mean()]
elif stat_measure is 'mode':
mode_ar = subset[feature].mode().values
if len(mode_ar) > 0:
row = [num_bin, mode_ar[0]]
else:
row = [num_bin, default_val]
else:
raise Exception('Unknown statistical measure: ' + stat_measure)
df_meancat.loc[len(df_meancat)] = row
for index, row_df in dataframe[dataframe[feature].isna()].iterrows():
for _, row_meancat in df_meancat.iterrows():
if row_df[bin_feature] in row_meancat['interval']:
new_dataframe.at[index, feature] = row_meancat['stat_measure']
return new_dataframe
def make_dummy_cols(dataframe, column, prefix, drop_dummy):
dummy = pd.get_dummies(dataframe[column], prefix=prefix)
dummy = dummy.drop(columns=prefix+'_'+drop_dummy)
dataframe = pd.concat([dataframe, dummy], axis=1)
dataframe = dataframe.drop(columns=column)
return dataframe
def cleaning(dataframe_raw):
dataframe = dataframe_raw.copy()
dataframe = dataframe.set_index('ID')
dataframe.loc[(dataframe['Age']<=13) & (dataframe['Education'].isna()), 'Education'] = 'Lower School/Kindergarten'
dataframe.loc[(dataframe['Age']==14) & (dataframe['Education'].isna()), 'Education'] = '8th Grade'
dataframe.loc[(dataframe['Age']<=17) & (dataframe['Education'].isna()), 'Education'] = '9 - 11th Grade'
dataframe.loc[(dataframe['Age']<=21) & (dataframe['Education'].isna()), 'Education'] = 'High School'
dataframe['Education'] = dataframe['Education'].fillna('Some College')
dataframe.loc[(dataframe['Age']<=20) & (dataframe['MaritalStatus'].isna()), 'MaritalStatus'] = 'NeverMarried'
dataframe.at[dataframe['MaritalStatus'].isna(), 'MaritalStatus'] = fill_bin_num(dataframe, 'MaritalStatus', 'Age', 5, 'mode',20)
dataframe = dataframe.drop(columns=['HHIncome'])
dataframe.loc[dataframe['HHIncomeMid'].isna(), 'HHIncomeMid'] = dataframe['HHIncomeMid'].mean()
dataframe.loc[dataframe['Poverty'].isna(), 'Poverty'] = dataframe['Poverty'].mean()
dataframe.loc[dataframe['HomeRooms'].isna(), 'HomeRooms'] = dataframe['HomeRooms'].mean()
dataframe.loc[dataframe['HomeOwn'].isna(), 'HomeOwn'] = dataframe['HomeOwn'].mode().values[0]
dataframe.loc[(dataframe['Work'].isna()) & (dataframe['Education'].isna()) & (dataframe['Age']<=20), 'Work'] = 'NotWorking'
dataframe.loc[dataframe['Work'].isna(), 'Work'] = dataframe['Work'].mode().values[0]
dataframe = fill_bin_num(dataframe, 'Weight', 'Age', 2, 'mean')
dataframe = dataframe.drop(columns=['HeadCirc'])
for index, row in dataframe.iterrows():
if np.isnan(row['Height']) and not np.isnan(row['Length']):
dataframe.at[index, 'Height'] = row['Length']
dataframe = fill_bin_num(dataframe, 'Height', 'Age', 2, 'mean')
dataframe = dataframe.drop(columns=['Length'])
for index, row in dataframe[dataframe['BMI'].isna()].iterrows():
dataframe.at[index, 'BMI'] = row['Weight'] / ((row['Height']/100)**2)
dataframe = dataframe.drop(columns='BMICatUnder20yrs')
dataframe = dataframe.drop(columns='BMI_WHO')
dataframe = fill_bin_num(dataframe, 'Pulse', 'Age', 10, 'mean')
dataframe.loc[(dataframe['Age']<10) & (dataframe['BPSysAve'].isna()), 'BPSysAve'] = 105
dataframe = fill_bin_num(dataframe, 'BPSysAve', 'Age', 5, 'mean', 10)
dataframe.loc[(dataframe['Age']<10) & (dataframe['BPDiaAve'].isna()), 'BPDiaAve'] = 60
dataframe = fill_bin_num(dataframe, 'BPDiaAve', 'Age', 5, 'mean', 10)
dataframe = dataframe.drop(columns='BPSys1')
dataframe = dataframe.drop(columns='BPDia1')
dataframe = dataframe.drop(columns='BPSys2')
dataframe = dataframe.drop(columns='BPDia2')
dataframe = dataframe.drop(columns='BPSys3')
dataframe = dataframe.drop(columns='BPDia3')
dataframe = dataframe.drop(columns=['Testosterone'])
dataframe.loc[(dataframe['Age']<10) & (dataframe['DirectChol'].isna()), 'DirectChol'] = 0
dataframe = fill_bin_num(dataframe, 'DirectChol', 'Age', 5, 'mean', 10)
dataframe.loc[(dataframe['Age']<10) & (dataframe['TotChol'].isna()), 'TotChol'] = 0
dataframe = fill_bin_num(dataframe, 'TotChol', 'Age', 5, 'mean', 10)
dataframe = dataframe.drop(columns=['UrineVol1'])
dataframe = dataframe.drop(columns=['UrineFlow1'])
dataframe = dataframe.drop(columns=['UrineVol2'])
dataframe = dataframe.drop(columns=['UrineFlow2'])
dataframe['Diabetes'] = dataframe['Diabetes'].fillna('No')
dataframe['DiabetesAge'] = dataframe['DiabetesAge'].fillna(0)
dataframe.loc[(dataframe['Age']<=12) & (dataframe['HealthGen'].isna()), 'HealthGen'] = 'Good'
dataframe = fill_bin_num(dataframe, 'HealthGen', 'Age', 5, 'mode', 10)
dataframe.loc[(dataframe['Age']<=12) & (dataframe['DaysMentHlthBad'].isna()), 'DaysMentHlthBad'] = 0
dataframe = fill_bin_num(dataframe, 'DaysMentHlthBad', 'Age', 5, 'mean', 10)
dataframe.loc[(dataframe['Age']<=15) & (dataframe['LittleInterest'].isna()), 'LittleInterest'] = 'None'
dataframe = fill_bin_num(dataframe, 'LittleInterest', 'Age', 5, 'mode', 15)
dataframe.loc[(dataframe['Age']<=12) & (dataframe['DaysMentHlthBad'].isna()), 'DaysMentHlthBad'] = 0
dataframe = fill_bin_num(dataframe, 'DaysMentHlthBad', 'Age', 5, 'mean', 10)
for index, row in dataframe.iterrows():
if np.isnan(row['nBabies']) and not np.isnan(row['nPregnancies']):
dataframe.at[index, 'nBabies'] = row['nPregnancies']
dataframe['nBabies'] = dataframe['nBabies'].fillna(0)
dataframe['nPregnancies'] = dataframe['nPregnancies'].fillna(0)
dataframe['Age1stBaby'] = dataframe['Age1stBaby'].fillna(0)
dataframe.loc[(dataframe['Age']==0) & (dataframe['SleepHrsNight'].isna()), 'SleepHrsNight'] = 14
dataframe.loc[(dataframe['Age']<=2) & (dataframe['SleepHrsNight'].isna()), 'SleepHrsNight'] = 12
dataframe.loc[(dataframe['Age']<=5) & (dataframe['SleepHrsNight'].isna()), 'SleepHrsNight'] = 10
dataframe.loc[(dataframe['Age']<=10) & (dataframe['SleepHrsNight'].isna()), 'SleepHrsNight'] = 9
dataframe.loc[(dataframe['Age']<=15) & (dataframe['SleepHrsNight'].isna()), 'SleepHrsNight'] = 8
dataframe['SleepHrsNight'] = dataframe['SleepHrsNight'].fillna(dataframe_raw['SleepHrsNight'].mean())
dataframe['SleepTrouble'] = dataframe['SleepTrouble'].fillna('No')
dataframe.loc[(dataframe['Age']<=4) & (dataframe['PhysActive'].isna()), 'PhysActive'] = 'No'
dataframe = fill_bin_num(dataframe, 'PhysActive', 'Age', 2, 'mode', 16)
dataframe['PhysActive'] = dataframe['PhysActive'].fillna('Yes') # Big assumption here. All kids between 4 and 16 are physically active
dataframe = dataframe.drop(columns=['PhysActiveDays'])
dataframe = dataframe.drop(columns=['TVHrsDay'])
dataframe = dataframe.drop(columns=['TVHrsDayChild'])
dataframe = dataframe.drop(columns=['CompHrsDay'])
dataframe = dataframe.drop(columns=['CompHrsDayChild'])
dataframe.loc[(dataframe['Age']<18) & (dataframe['Alcohol12PlusYr'].isna()), 'Alcohol12PlusYr'] = 'No'
dataframe = fill_bin_num(dataframe, 'Alcohol12PlusYr', 'Age', 5, 'mode', 18)
dataframe.loc[(dataframe['Age']<18) & (dataframe['AlcoholDay'].isna()), 'AlcoholDay'] = 0
dataframe = fill_bin_num(dataframe, 'AlcoholDay', 'Age', 5, 'mean', 18)
dataframe.loc[(dataframe['Age']<18) & (dataframe['AlcoholYear'].isna()), 'AlcoholYear'] = 0
dataframe = fill_bin_num(dataframe, 'AlcoholYear', 'Age', 5, 'mean', 18)
dataframe.loc[(dataframe['Age']<20) & (dataframe['SmokeNow'].isna()), 'SmokeNow'] = 'No'
dataframe = fill_bin_num(dataframe, 'SmokeNow', 'Age', 5, 'mode', 20)
dataframe['Smoke100'] = dataframe['Smoke100'].fillna('No')
dataframe['Smoke100n'] = dataframe['Smoke100n'].fillna('No')
dataframe.loc[(dataframe['SmokeNow']=='No') & (dataframe['SmokeAge'].isna()), 'SmokeAge'] = 0
dataframe = fill_bin_num(dataframe, 'SmokeAge', 'Age', 5, 'mean', 20)
dataframe.loc[(dataframe['Age']<18) & (dataframe['Marijuana'].isna()), 'Marijuana'] = 'No'
dataframe.loc[(dataframe['Marijuana'].isna()) & (dataframe['SmokeNow']=='No'), 'Marijuana'] = 'No'
dataframe = fill_bin_num(dataframe, 'Marijuana', 'Age', 5, 'mode', 20)
dataframe.loc[(dataframe['Marijuana']=='No') & (dataframe['AgeFirstMarij'].isna()), 'AgeFirstMarij'] = 0
dataframe = fill_bin_num(dataframe, 'AgeFirstMarij', 'Age', 5, 'mean', 20)
dataframe.loc[(dataframe['Marijuana']=='No') & (dataframe['RegularMarij'].isna()), 'RegularMarij'] = 'No'
dataframe = fill_bin_num(dataframe, 'RegularMarij', 'Age', 5, 'mode', 20)
dataframe.loc[(dataframe['RegularMarij']=='No') & (dataframe['AgeRegMarij'].isna()), 'AgeRegMarij'] = 0
dataframe = fill_bin_num(dataframe, 'AgeRegMarij', 'Age', 5, 'mean', 20)
dataframe.loc[(dataframe['Age']<18) & (dataframe['HardDrugs'].isna()), 'HardDrugs'] = 'No'
dataframe = fill_bin_num(dataframe, 'HardDrugs', 'Age', 5, 'mode', 18)
mode_sex_age = dataframe['SexAge'].mode()[0]
dataframe.loc[(dataframe['Age']<=mode_sex_age) & (dataframe['SexEver'].isna()), 'SexEver'] = 'No'
dataframe['SexEver'] = dataframe['SexEver'].fillna('Yes')
dataframe.loc[(dataframe['SexEver']=='No') & (dataframe['SexAge'].isna()), 'SexAge'] = 0
dataframe.loc[(dataframe['SexAge'].isna() & (dataframe['Age']<mode_sex_age)), 'SexAge'] = dataframe.loc[(dataframe['SexAge'].isna() & (dataframe['Age']<mode_sex_age)), 'Age']
dataframe['SexAge'] = dataframe['SexAge'].fillna(mode_sex_age)
dataframe.loc[(dataframe['SexEver']=='No') & (dataframe['SexNumPartnLife'].isna()), 'SexNumPartnLife'] = 0
dataframe = fill_bin_num(dataframe, 'SexNumPartnLife', 'Age', 5, 'mean')
dataframe['SexNumPartnLife'] = dataframe_raw.loc[(dataframe_raw['Age'] >= 60) & (dataframe_raw['Age'] <= 70), 'SexNumPartnLife'].mode()[0] # Missing values for the elderly. Assumed that lifetime sex partners do not increase after 60.
dataframe.loc[(dataframe['SexEver']=='No') & (dataframe['SexNumPartYear'].isna()), 'SexNumPartYear'] = 0
dataframe = fill_bin_num(dataframe, 'SexNumPartYear', 'Age', 10, 'mean')
dataframe['SexNumPartYear'] = dataframe['SexNumPartYear'].fillna(0)
dataframe = dataframe.drop(columns=['SameSex'])
dataframe = dataframe.drop(columns=['SexOrientation'])
dataframe['PregnantNow'] = dataframe['PregnantNow'].fillna('No')
# Making dummy variables
dataframe['male'] = 1*(dataframe['Gender'] == 'male')
dataframe = dataframe.drop(columns=['Gender'])
dataframe['white'] = np.where(dataframe['Race1'] == 'white',1,0)
dataframe = dataframe.drop(columns=['Race1'])
dataframe = make_dummy_cols(dataframe, 'Education', 'education', '8th Grade')
dataframe = make_dummy_cols(dataframe, 'MaritalStatus', 'maritalstatus', 'Separated')
dataframe = make_dummy_cols(dataframe, 'HomeOwn', 'homeown', 'Other')
dataframe = make_dummy_cols(dataframe, 'Work', 'work', 'Looking')
dataframe['Diabetes'] = np.where(dataframe['Diabetes'] == 'Yes',1,0)
dataframe = make_dummy_cols(dataframe, 'HealthGen', 'healthgen', 'Poor')
dataframe = make_dummy_cols(dataframe, 'LittleInterest', 'littleinterest', 'None')
dataframe = make_dummy_cols(dataframe, 'Depressed', 'depressed', 'None')
dataframe['SleepTrouble'] = np.where(dataframe['SleepTrouble'] == 'Yes',1,0)
dataframe['PhysActive'] = np.where(dataframe['PhysActive'] == 'Yes',1,0)
dataframe['Alcohol12PlusYr'] = np.where(dataframe['Alcohol12PlusYr'] == 'Yes',1,0)
dataframe['SmokeNow'] = np.where(dataframe['SmokeNow'] == 'Yes',1,0)
dataframe['Smoke100'] = np.where(dataframe['Smoke100'] == 'Yes',1,0)
dataframe['Smoke100n'] = np.where(dataframe['Smoke100n'] == 'Yes',1,0)
dataframe['Marijuana'] = np.where(dataframe['Marijuana'] == 'Yes',1,0)
dataframe['RegularMarij'] = np.where(dataframe['RegularMarij'] == 'Yes',1,0)
dataframe['HardDrugs'] = np.where(dataframe['HardDrugs'] == 'Yes',1,0)
dataframe['SexEver'] = np.where(dataframe['SexEver'] == 'Yes',1,0)
dataframe['PregnantNow'] = np.where(dataframe['PregnantNow'] == 'Yes',1,0)
return dataframe