-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathextractor.py
228 lines (179 loc) · 8.64 KB
/
extractor.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
import pandas as pd
from datetime import datetime
date_converter = lambda x: datetime.strptime(x[0:19], '%Y-%m-%d %H:%M:%S')
def date_subtractor(x, y):
return (x - y).days
def get_table_mimic(padj, table_name):
path_mimicroot = ""
return path_mimicroot + padj + "/" + "{0}/{1}-{0}.txt".format(pid, table_name)
def filler(x):
return x
'''
Physiological variables init
'''
class PhysioVar:
"""A physiological variable class"""
def __init__(self, id, min, max, unit, name):
self.id = id
self.min = min
self.max = max
self.name = name
vararr = []
vararr.append(PhysioVar(618, 0, 250, 'bpm', 'RESP'))
vararr.append(PhysioVar(211, 0, 250, 'bpm', 'HEART'))
vararr.append(PhysioVar(676, 25, 42, 'c', 'TEMP'))
##null
#vararr.append(PhysioVar(3834, 400, 5000, '1000 cells per micro L', 'WBC'))
vararr.append(PhysioVar(1162, 4, 500, 'mg per dL', 'BUN'))
vararr.append(PhysioVar(791, 0.1, 9, 'mg per dL', 'CREAT'))
#vararr.append(PhysioVar(2633, 90, 180, 'mmHg', 'SYSBP'))
#vararr.append(PhysioVar(2632, 60, 110, 'mmHg', 'DIABP'))
vararr.append(PhysioVar(52, 70, 110, 'mmHg', 'MEANBP'))
vararr.append(PhysioVar(646, 60, 100, '%', 'O2SAT'))
vararr.append(PhysioVar(818, 0, 10, 'mg per dL', 'LACTIC'))
#vararr.append(PhysioVar(6256, 3000, 1000000, 'cells per L', 'PLAT'))
#vararr.append(PhysioVar(6256, -1, 99999999999, 'cells per L', 'PLAT'))
#vararr.append(PhysioVar(833, 2000, 8000, 'mg per dL', 'RBC'))
vararr.append(PhysioVar(813, 19, 60, '%', 'HEMATOCRIT'))
vararr.append(PhysioVar(837, 120, 160, 'mEq per dL', 'SODIUM'))
vararr.append(PhysioVar(829, 2.2, 8, 'mEq per dL', 'POTASSIUM'))
vararr.append(PhysioVar(786, 4.8, 12, 'mg per dL', 'CALCIUM'))
vararr.append(PhysioVar(821, 0, 10, 'mg per dL', 'MAGNESIUM'))
vararr.append(PhysioVar(772, 0.5, 18, 'mg per dL', 'ALBUMIN'))
vararr.append(PhysioVar(780, 6.8, 7.8, ' ', 'ARTERIALPH'))
#vararr.append(PhysioVar(3750, 0, 1000, ' ', 'URINEOUTFOLEY'))
vararr.append(PhysioVar(763, 20, 200, 'kg', 'WEIGHT'))
#vararr.append(PhysioVar(3750, 6.8, 7.8, ' ', 'MECHANICALVENTILATION'))
for i in range(len(vararr)):
print vararr[i].id
'''
End of Physiological variables
'''
dfs = []
for j in range (0, 10):
# padj: leftmost 2 digits
padj = str(j).rjust(2, '0')
for i in range(1, 20):
try:
# padj: rightmost 3 digits
padi = str(i).rjust(3, '0')
pid = padj + padi
file_dpatients = padj + "/" + "{0}/D_PATIENTS-{0}.txt".format(pid)
#file_icustayev = "{0}/ICUSTAYEVENTS-{0}.txt".format(t)
file_icustaydet = padj + "/" + "{0}/ICUSTAY_DETAIL-{0}.txt".format(pid)
data_tmp = pd.read_csv(file_dpatients, header = 0, parse_dates=True, usecols=['SUBJECT_ID', 'HOSPITAL_EXPIRE_FLG'])
#data_tmp_icustays = pd.read_csv(file_icustayev, header = 0, parse_dates=True, usecols=['SUBJECT_ID', 'ICUSTAY_ID','INTIME'])
data_tmp_icustaydet = pd.read_csv(file_icustaydet, header = 0, parse_dates=True, usecols=['SUBJECT_ID', 'ICUSTAY_ID', 'ICUSTAY_ADMIT_AGE', 'ICUSTAY_INTIME','ICUSTAY_OUTTIME'])
df = pd.merge(data_tmp, data_tmp_icustaydet, on = 'SUBJECT_ID')
df['ICUSTAY_INTIME'] = df['ICUSTAY_INTIME'].apply(date_converter)
df['ICUSTAY_OUTTIME'] = df['ICUSTAY_OUTTIME'].apply(date_converter)
try:
df['STAY'] = df.apply(lambda x: date_subtractor(x['ICUSTAY_OUTTIME'], x['ICUSTAY_INTIME']), axis = 1)
#df['RESP'] = df.apply(lambda x: date_subtractor(x['ICUSTAY_OUTTIME'], x['ICUSTAY_INTIME']), axis = 1)
except Exception, e2:
print e2
data_merged = df.query('ICUSTAY_ADMIT_AGE>15 and STAY>0')
#data_merged = data_merged.merge(TABLE_CHARTEV,on=['ICUSTAY_ID'], how='inner').groupby(['ICUSTAY_ID'], as_index=False)['VALUE1NUM'].mean()
'''
accumulate only resp variable
TABLE_CHARTEV = TABLE_CHARTEV.query('ITEMID == {0} and VALUE1NUM >= {1} and VALUE1NUM <= {2}'.format(physivar.id, physivar.min, physivar.max ) )
#print TABLE_CHARTEV.groupby('ICUSTAY_ID')['VALUE1NUM'].mean()
aggregated = TABLE_CHARTEV.groupby('ICUSTAY_ID').mean()['VALUE1NUM']
aggregated.name = 'RESP'
#print "***********************$$$$"
#print aggregated
#print "***********************"
data_merged = data_merged.join(aggregated,on='ICUSTAY_ID')
'''
TABLE_CHARTEV = pd.read_csv(get_table_mimic(padj, 'CHARTEVENTS'), header = 0, usecols=['SUBJECT_ID', 'ITEMID', 'ICUSTAY_ID', 'VALUE1NUM' ])
for physivar in vararr:
aggregated = TABLE_CHARTEV.query('ITEMID == {0} and VALUE1NUM >= {1} and VALUE1NUM <= {2}'.format(physivar.id, physivar.min, physivar.max ) )
aggregated = aggregated.groupby('ICUSTAY_ID').mean()['VALUE1NUM']
aggregated.name = physivar.name
#print "***********************$$$$"
#print aggregated
#print "***********************"
data_merged = data_merged.join(aggregated, on='ICUSTAY_ID')
del df
dfs.append(data_merged)
except Exception, e:
print e
del data_tmp
#del data_tmp_icustays
del data_merged
data = pd.concat(dfs)
#del data['ICUSTAY_OUTTIME']
#del data['ICUSTAY_INTIME']
print data
print "Total data: {0}".format(data.shape[0])
data = data.query('HOSPITAL_EXPIRE_FLG == \'N\'')
print "Total data, after expire: {0}".format(data.shape[0])
data.to_csv('all.csv')
no_read = data.groupby(['SUBJECT_ID'])
no_read = no_read.filter(lambda x: len(x) < 2)
no_read['CLASS'] = no_read.apply(lambda x: filler(0), axis = 1)
#no_read = data[data.groupby('ICUSTAY_ID').pid.transform(len) > 1]
print no_read
print "Total not readmit, {0}".format(no_read.shape[0])
no_read.to_csv('negative.csv')
#readmitted = data.groupby(['SUBJECT_ID'])
#data['ICUSTAY_INTIME'] = data.apply(lambda x: str(x['ICUSTAY_INTIME']), axis = 1)
#readmitted = data.groupby('SUBJECT_ID').min()['ICUSTAY_ID']
readmitted = data.groupby(['SUBJECT_ID']).filter(lambda x: len(x) > 1)
print readmitted
readmitted = readmitted.sort_values('ICUSTAY_ID', ascending=True).drop_duplicates(['SUBJECT_ID'])
readmitted['CLASS'] = readmitted.apply(lambda x: filler(1), axis = 1)
#readmitted = data.join(readmitted, on='ICUSTAY_ID')
print readmitted
readmitted.to_csv('positive.csv')
mimicall = []
mimicall.append(no_read)
mimicall.append(readmitted)
finaldata = pd.concat(mimicall)
finaldata.drop(finaldata.columns[[0, 1, 2, 3, 4, 5, 6]], axis=1, inplace=True) # df.columns is zero-based pd.Index
#finaldata.drop(['ICUSTAY_OUTTIME', 'ICUSTAY_INTIME', 'HOSPITAL_EXPIRE_FLG', 'ICUSTAY_ID', 'SUBJECT_ID', 'STAY' ], axis=1, inplace=True)
finaldata.to_csv('finaldata.csv', index = False)
'''
### HELPING CODES
## Own date parser
dateparser = lambda x: pd.datetime.strptime(x, '%Y-%m-%d %H:%M:%S')
# Which makes your read command:
pd.read_csv(infile, parse_dates=['columnName'], date_parser=dateparse)
# Or combine two columns into a single DateTime column
pd.read_csv(infile, parse_dates={'datetime': ['date', 'time']}, date_parser=dateparse)
#### HOW TO READ ALL FILES
#files = glob.glob('*.csv')
#weather_dfs = [pd.read_csv(fp, names=columns) for fp in files]
#weather = pd.concat(weather_dfs)
#### HOW TO INDEX
print df.iloc[0, 2].month()
subset1 = df.iloc[:, :]
print df
#### HOW TO QUERY DATA
#data_merged = df.query('ICUSTAY_ADMIT_AGE>0').query('0<b<2')
#df_filtered = df.query('a>0 and 0<b<2')
### HOW TO WRITE DATETIME DATE PARSER and DAY SUBTRACT
datetime_object = datetime.strptime('2896-10-24 11:42:00', '%Y-%m-%d %H:%M:%S')
datetime_object2 = datetime.strptime('2846-10-24 11:42:00', '%Y-%m-%d %H:%M:%S')
print round((datetime_object - datetime_object2).days/365.2, 2)
'''
'''
#### HOW TO DO PIVOT
dfs = []
path = 'D:\\'
# data = pd.DataFrame()
for file in raw_files:
data_tmp = pd.read_csv(path + file, engine='c',
compression='gzip',
low_memory=False,
usecols=['date', 'Value', 'ID'])
data_tmp = data_tmp.pivot(index='date', columns='ID',
values='Value')
dfs.append(data_tmp)
del data_tmp
data = pd.concat(dfs)
'''
'''
#### HOW TO WRITE LAMBDA
#doubler = lambda x: x*2
'''