-
Notifications
You must be signed in to change notification settings - Fork 0
/
Index_Prediction.py
184 lines (147 loc) · 7.41 KB
/
Index_Prediction.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
import pandas as pd
import numpy as np
import datetime
from sklearn import linear_model
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import r2_score
from sklearn.model_selection import train_test_split
class HD_index_prediction:
def __init__(self):
pass
def load_data(self,info_data,X_data,Y_data):
info_data[['HOSPITAL_ID', 'PATIENT_NK']] = info_data[['HOSPITAL_ID', 'PATIENT_NK']].astype(str)
#info_data=info_data[info_data['AGE']>18]
X_data[['HOSPITAL_ID', 'PATIENT_NK']] = X_data[['HOSPITAL_ID', 'PATIENT_NK']].astype(str)
self.X_data=X_data
self.Y_data=Y_data
self.info_data=info_data
def preprocess_training(self,index):
self.index=index
time_limit=60
delta_time=datetime.timedelta(days=time_limit)
X_list=[]
Y_list=[]
patient_list=[]
index_average_value_list=[]
for i in range(len(self.Y_data)):
if i%1000==0:
print('加载数据',str(round(i/len(self.Y_data)*100,2))+'%')
if eval(Y_data[index].iloc[i])==[]:
continue
p=self.Y_data['PATIENT_NK'].iloc[i]
h=self.Y_data['HOSPITAL_ID'].iloc[i]
each_X=self.X_data[(self.X_data['PATIENT_NK']==p)&(self.X_data['HOSPITAL_ID']==h)]
print(each_X)
if len(each_X)==0:
continue
each_info=self.info_data[(self.info_data['PATIENT_NK']==p)&(self.info_data['HOSPITAL_ID']==h)]
if len(each_info)==0:
continue
age=each_info['AGE'].iloc[0]
gender=each_info['GENDER'].iloc[0]
if gender==1:
gender=1
else:
gender=0
HD_time_list=list(each_X['CREATE_TIME'])
time_list=eval(self.Y_data[index+'日期'].iloc[i])
if len(time_list)<5:
continue
time_matrix=np.zeros((len(time_list)-4,len(HD_time_list)))
for l in range(4,len(time_list)):
for j in range(len(HD_time_list)):
HD_time=datetime.datetime.strptime(HD_time_list[j], "%Y-%m-%d %H.%M.%S")
index_time=datetime.datetime.strptime(time_list[l], "%Y-%m-%d %H.%M.%S")
value=max((index_time-HD_time).days,0)
if value==0:
value=999999#这样就取不到这个时间点
time_matrix[l-4,j]=value
HD_index_list=[]
j_list=[]
for l in range(len(time_list)-4):
if np.min(time_matrix[l])<time_limit:
j=np.where(time_matrix[l]==np.min(time_matrix[l]))[0][0]
if j in j_list:
continue
j_list.append(j)
HD_index_list.append([l+4,j])
data_list=eval(self.Y_data[index].iloc[i])
for lj in HD_index_list:
l=lj[0]
j=lj[1]
HD_time=each_X['CREATE_TIME'].iloc[j]
HD_time=datetime.datetime.strptime(HD_time, "%Y-%m-%d %H.%M.%S")
index_time=time_list[l]
index_time=datetime.datetime.strptime(index_time, "%Y-%m-%d %H.%M.%S")
if index_time>HD_time and index_time-HD_time<delta_time:
each_x=[]
each_x=each_x
if each_X['NEOPATHY_TYPE'].iloc[j]!=0:
LBP=1
else:
LBP=0
DBP_list=eval(each_X['DBP'].iloc[j].replace('nan','0'))
MEAN_AP_list=eval(each_X['MEAN_AP'].iloc[j].replace('nan','0'))
last_SBP_list=eval(each_X['SBP'].iloc[j].replace('nan','0'))
last_SBP_list=np.array(last_SBP_list)
last_SBP_list=last_SBP_list[last_SBP_list>0]
DBP_list=np.array(DBP_list)
DBP_list=DBP_list[DBP_list>0]
if len(last_SBP_list)<2 or len(DBP_list)<2:
continue
SBP_mean=np.mean(last_SBP_list)
SBP_std=np.std(last_SBP_list)
SBP_diff=last_SBP_list[:-1]-last_SBP_list[1:]
SBP_diff_abs_mean = np.mean(np.abs(SBP_diff))
SBP_diff_abs_std= np.std(np.abs(SBP_diff))
SBP_diff_mean = np.mean(SBP_diff)
SBP_diff_std = np.std(SBP_diff)
DBP_mean=np.mean(DBP_list)
DBP_std=np.std(DBP_list)
DBP_diff=DBP_list[:-1]-DBP_list[1:]
DBP_diff_abs_mean = np.mean(np.abs(DBP_diff))
DBP_diff_abs_std= np.std(np.abs(DBP_diff))
DBP_diff_mean = np.mean(DBP_diff)
DBP_diff_std = np.std(DBP_diff)
MEAN_AP_list=np.array(MEAN_AP_list)
MEAN_AP_list=MEAN_AP_list[MEAN_AP_list>0]
MEAN_AP_mean=np.mean(MEAN_AP_list)
MEAN_AP_std=np.std(MEAN_AP_list)
each_x=each_x+[SBP_mean,SBP_std,
SBP_diff_abs_mean,SBP_diff_abs_std,
SBP_diff_mean,SBP_diff_std,
DBP_mean,DBP_std,
DBP_diff_abs_mean,DBP_diff_abs_std,
DBP_diff_mean,DBP_diff_std,
MEAN_AP_mean,MEAN_AP_std,
age,data_list[l-4],data_list[l-3],data_list[l-2],data_list[l-1]]
each_x=each_x + [gender,LBP]
if data_list[l]<=0:
continue
X_list.append(each_x)
Y_list.append(data_list[l])
if [h,p] not in patient_list:
index_average_value_list.append(np.mean(data_list))
patient_list.append([h,p])
X_list=np.array(X_list)
Y_list=np.array(Y_list)
stdsc=StandardScaler().fit(X_list[:,:-2])
np.savetxt('数据基础：特征值'+str(index)+'.csv', X_list, delimiter=",")
np.savetxt('数据基础：目标值'+str(index)+'.csv', Y_list, delimiter=",")
X_std=stdsc.fit_transform(X_list[:,:-2])#正态分布标准化
X_std = np.concatenate((X_std,X_list[:,-2:]),axis=1)
X_train,X_test,y_train,y_test= train_test_split(X_std,Y_list,test_size=0.2,random_state=0,shuffle=True)
#X_train,X_val,y_train,y_val = train_test_split(X_train,y_train,test_size=0.25,random_state=0,shuffle=True)
lr = linear_model.LassoLarsCV(cv=2,n_jobs=-1)
lr.fit(X_train, y_train)
lr.score(X_train,y_train)
# score = r2_score(y_train, lr.predict(X_train)),
self.model=lr
self.stdsc=stdsc
self.score=score
return X_list,Y_list,score
def predict(self,x):
X_std=self.stdsc.fit_transform(x[:,:-2])#正态分布标准化
X_std = np.concatenate((X_std,x[:,-2:]),axis=1)
y_pred = self.model.predict(X_std)
return y_pred