thbndi commited on
Commit
dd7c0c8
1 Parent(s): 31f2359

Update dataset_utils.py

Browse files
Files changed (1) hide show
  1. dataset_utils.py +28 -16
dataset_utils.py CHANGED
@@ -77,7 +77,7 @@ def vocab(task,diag_flag,proc_flag,out_flag,chart_flag,med_flag,lab_flag):
77
 
78
  return len(condVocabDict),len(procVocabDict),len(medVocabDict),len(outVocabDict),len(chartVocabDict),len(labVocabDict),ethVocabDict,genderVocabDict,ageVocabDict,insVocabDict
79
 
80
- def concat_data(data,task,feat_cond,feat_proc,feat_out,feat_chart,feat_meds,feat_lab,outDict,chartDict,condDict,procDict,medDict):
81
  meds=data['Med']
82
  proc = data['Proc']
83
  out = data['Out']
@@ -98,7 +98,9 @@ def concat_data(data,task,feat_cond,feat_proc,feat_out,feat_chart,feat_meds,feat
98
  ##########COND#########
99
  if (feat_cond):
100
  #get all conds
101
- conds=pd.DataFrame(condDict,columns=['COND'])
 
 
102
  features=pd.DataFrame(np.zeros([1,len(conds)]),columns=conds['COND'])
103
 
104
  #onehot encode
@@ -120,11 +122,13 @@ def concat_data(data,task,feat_cond,feat_proc,feat_out,feat_chart,feat_meds,feat
120
 
121
  ##########PROC#########
122
  if (feat_proc):
 
 
123
 
124
  if proc :
125
  feat=proc.keys()
126
  proc_val=[proc[key] for key in feat]
127
- procedures=pd.DataFrame(procDict,columns=['PROC'])
128
  features=pd.DataFrame(np.zeros([1,len(procedures)]),columns=procedures['PROC'])
129
  features.columns=pd.MultiIndex.from_product([["PROC"], features.columns])
130
  procs=pd.DataFrame(columns=feat)
@@ -140,11 +144,13 @@ def concat_data(data,task,feat_cond,feat_proc,feat_out,feat_chart,feat_meds,feat
140
 
141
  ##########OUT#########
142
  if (feat_out):
 
 
143
 
144
  if out :
145
  feat=out.keys()
146
  out_val=[out[key] for key in feat]
147
- outputs=pd.DataFrame(outDict,columns=['OUT'])
148
  features=pd.DataFrame(np.zeros([1,len(outputs)]),columns=outputs['OUT'])
149
  features.columns=pd.MultiIndex.from_product([["OUT"], features.columns])
150
  outs=pd.DataFrame(columns=feat)
@@ -153,60 +159,68 @@ def concat_data(data,task,feat_cond,feat_proc,feat_out,feat_chart,feat_meds,feat
153
  outs.columns=pd.MultiIndex.from_product([["OUT"], outs.columns])
154
  out_df = pd.concat([features,outs],ignore_index=True).fillna(0)
155
  else:
156
- outputs=pd.DataFrame(outDict,columns=['OUT'])
157
  features=pd.DataFrame(np.zeros([1,len(outputs)]),columns=outputs['OUT'])
158
  features.columns=pd.MultiIndex.from_product([["OUT"], features.columns])
159
  out_df=features.fillna(0)
160
 
161
  ##########CHART#########
162
  if (feat_chart):
 
 
 
163
  if chart:
164
  charts=chart['val']
165
  feat=charts.keys()
166
  chart_val=[charts[key] for key in feat]
167
- charts=pd.DataFrame(chartDict,columns=['CHART'])
168
  features=pd.DataFrame(np.zeros([1,len(charts)]),columns=charts['CHART'])
169
  features.columns=pd.MultiIndex.from_product([["CHART"], features.columns])
 
170
  chart=pd.DataFrame(columns=feat)
171
  for c,v in zip(feat,chart_val):
172
  chart[c]=v
173
  chart.columns=pd.MultiIndex.from_product([["CHART"], chart.columns])
174
  chart_df = pd.concat([features,chart],ignore_index=True).fillna(0)
175
-
176
  else:
177
- charts=pd.DataFrame(chartDict,columns=['CHART'])
178
  features=pd.DataFrame(np.zeros([1,len(charts)]),columns=charts['CHART'])
179
  features.columns=pd.MultiIndex.from_product([["CHART"], features.columns])
180
  chart_df=features.fillna(0)
181
 
182
  ##########LAB#########
183
  if (feat_lab):
 
 
 
184
  if chart:
185
  charts=chart['val']
186
  feat=charts.keys()
187
  chart_val=[charts[key] for key in feat]
188
- charts=pd.DataFrame(chartDict,columns=['LAB'])
189
  features=pd.DataFrame(np.zeros([1,len(charts)]),columns=charts['LAB'])
190
  features.columns=pd.MultiIndex.from_product([["LAB"], features.columns])
191
 
192
-
193
  chart=pd.DataFrame(columns=feat)
194
  for c,v in zip(feat,chart_val):
195
  chart[c]=v
196
  chart.columns=pd.MultiIndex.from_product([["LAB"], chart.columns])
197
  chart_df = pd.concat([features,chart],ignore_index=True).fillna(0)
198
  else:
199
- charts=pd.DataFrame(chartDict,columns=['LAB'])
200
  features=pd.DataFrame(np.zeros([1,len(charts)]),columns=charts['LAB'])
201
  features.columns=pd.MultiIndex.from_product([["LAB"], features.columns])
202
  chart_df=features.fillna(0)
203
-
204
  ###MEDS
205
  if (feat_meds):
 
 
 
206
  if meds:
207
  feat=meds['signal'].keys()
208
  med_val=[meds['amount'][key] for key in feat]
209
- meds=pd.DataFrame(medDict,columns=['MEDS'])
210
  features=pd.DataFrame(np.zeros([1,len(meds)]),columns=meds['MEDS'])
211
  features.columns=pd.MultiIndex.from_product([["MEDS"], features.columns])
212
 
@@ -216,7 +230,7 @@ def concat_data(data,task,feat_cond,feat_proc,feat_out,feat_chart,feat_meds,feat
216
  med.columns=pd.MultiIndex.from_product([["MEDS"], med.columns])
217
  meds_df = pd.concat([features,med],ignore_index=True).fillna(0)
218
  else:
219
- meds=pd.DataFrame(medDict,columns=['MEDS'])
220
  features=pd.DataFrame(np.zeros([1,len(meds)]),columns=meds['MEDS'])
221
  features.columns=pd.MultiIndex.from_product([["MEDS"], features.columns])
222
  meds_df=features.fillna(0)
@@ -330,8 +344,6 @@ def generate_ml(dyn,stat,demo,concat_cols,concat):
330
  else:
331
  dyn_df=pd.concat([dyn_df,agg],axis=0)
332
  dyn_df=dyn_df.T
333
- print(dyn_df.columns)
334
- print(dyn_df)
335
  dyn_df.columns = dyn_df.iloc[0]
336
  dyn_df=dyn_df.iloc[1:,:]
337
 
 
77
 
78
  return len(condVocabDict),len(procVocabDict),len(medVocabDict),len(outVocabDict),len(chartVocabDict),len(labVocabDict),ethVocabDict,genderVocabDict,ageVocabDict,insVocabDict
79
 
80
+ def concat_data(data,task,feat_cond,feat_proc,feat_out,feat_chart,feat_meds,feat_lab):
81
  meds=data['Med']
82
  proc = data['Proc']
83
  out = data['Out']
 
98
  ##########COND#########
99
  if (feat_cond):
100
  #get all conds
101
+ with open("./data/dict/"+task+"/condVocab", 'rb') as fp:
102
+ conDict = pickle.load(fp)
103
+ conds=pd.DataFrame(conDict,columns=['COND'])
104
  features=pd.DataFrame(np.zeros([1,len(conds)]),columns=conds['COND'])
105
 
106
  #onehot encode
 
122
 
123
  ##########PROC#########
124
  if (feat_proc):
125
+ with open("./data/dict/"+task+"/procVocab", 'rb') as fp:
126
+ procDic = pickle.load(fp)
127
 
128
  if proc :
129
  feat=proc.keys()
130
  proc_val=[proc[key] for key in feat]
131
+ procedures=pd.DataFrame(procDic,columns=['PROC'])
132
  features=pd.DataFrame(np.zeros([1,len(procedures)]),columns=procedures['PROC'])
133
  features.columns=pd.MultiIndex.from_product([["PROC"], features.columns])
134
  procs=pd.DataFrame(columns=feat)
 
144
 
145
  ##########OUT#########
146
  if (feat_out):
147
+ with open("./data/dict/"+task+"/outVocab", 'rb') as fp:
148
+ outDic = pickle.load(fp)
149
 
150
  if out :
151
  feat=out.keys()
152
  out_val=[out[key] for key in feat]
153
+ outputs=pd.DataFrame(outDic,columns=['OUT'])
154
  features=pd.DataFrame(np.zeros([1,len(outputs)]),columns=outputs['OUT'])
155
  features.columns=pd.MultiIndex.from_product([["OUT"], features.columns])
156
  outs=pd.DataFrame(columns=feat)
 
159
  outs.columns=pd.MultiIndex.from_product([["OUT"], outs.columns])
160
  out_df = pd.concat([features,outs],ignore_index=True).fillna(0)
161
  else:
162
+ outputs=pd.DataFrame(outDic,columns=['OUT'])
163
  features=pd.DataFrame(np.zeros([1,len(outputs)]),columns=outputs['OUT'])
164
  features.columns=pd.MultiIndex.from_product([["OUT"], features.columns])
165
  out_df=features.fillna(0)
166
 
167
  ##########CHART#########
168
  if (feat_chart):
169
+ with open("./data/dict/"+task+"/chartVocab", 'rb') as fp:
170
+ chartDic = pickle.load(fp)
171
+
172
  if chart:
173
  charts=chart['val']
174
  feat=charts.keys()
175
  chart_val=[charts[key] for key in feat]
176
+ charts=pd.DataFrame(chartDic,columns=['CHART'])
177
  features=pd.DataFrame(np.zeros([1,len(charts)]),columns=charts['CHART'])
178
  features.columns=pd.MultiIndex.from_product([["CHART"], features.columns])
179
+
180
  chart=pd.DataFrame(columns=feat)
181
  for c,v in zip(feat,chart_val):
182
  chart[c]=v
183
  chart.columns=pd.MultiIndex.from_product([["CHART"], chart.columns])
184
  chart_df = pd.concat([features,chart],ignore_index=True).fillna(0)
 
185
  else:
186
+ charts=pd.DataFrame(chartDic,columns=['CHART'])
187
  features=pd.DataFrame(np.zeros([1,len(charts)]),columns=charts['CHART'])
188
  features.columns=pd.MultiIndex.from_product([["CHART"], features.columns])
189
  chart_df=features.fillna(0)
190
 
191
  ##########LAB#########
192
  if (feat_lab):
193
+ with open("./data/dict/"+task+"/labsVocab", 'rb') as fp:
194
+ chartDic = pickle.load(fp)
195
+
196
  if chart:
197
  charts=chart['val']
198
  feat=charts.keys()
199
  chart_val=[charts[key] for key in feat]
200
+ charts=pd.DataFrame(chartDic,columns=['LAB'])
201
  features=pd.DataFrame(np.zeros([1,len(charts)]),columns=charts['LAB'])
202
  features.columns=pd.MultiIndex.from_product([["LAB"], features.columns])
203
 
 
204
  chart=pd.DataFrame(columns=feat)
205
  for c,v in zip(feat,chart_val):
206
  chart[c]=v
207
  chart.columns=pd.MultiIndex.from_product([["LAB"], chart.columns])
208
  chart_df = pd.concat([features,chart],ignore_index=True).fillna(0)
209
  else:
210
+ charts=pd.DataFrame(chartDic,columns=['LAB'])
211
  features=pd.DataFrame(np.zeros([1,len(charts)]),columns=charts['LAB'])
212
  features.columns=pd.MultiIndex.from_product([["LAB"], features.columns])
213
  chart_df=features.fillna(0)
214
+
215
  ###MEDS
216
  if (feat_meds):
217
+ with open("./data/dict/"+task+"/medVocab", 'rb') as fp:
218
+ medDic = pickle.load(fp)
219
+
220
  if meds:
221
  feat=meds['signal'].keys()
222
  med_val=[meds['amount'][key] for key in feat]
223
+ meds=pd.DataFrame(medDic,columns=['MEDS'])
224
  features=pd.DataFrame(np.zeros([1,len(meds)]),columns=meds['MEDS'])
225
  features.columns=pd.MultiIndex.from_product([["MEDS"], features.columns])
226
 
 
230
  med.columns=pd.MultiIndex.from_product([["MEDS"], med.columns])
231
  meds_df = pd.concat([features,med],ignore_index=True).fillna(0)
232
  else:
233
+ meds=pd.DataFrame(medDic,columns=['MEDS'])
234
  features=pd.DataFrame(np.zeros([1,len(meds)]),columns=meds['MEDS'])
235
  features.columns=pd.MultiIndex.from_product([["MEDS"], features.columns])
236
  meds_df=features.fillna(0)
 
344
  else:
345
  dyn_df=pd.concat([dyn_df,agg],axis=0)
346
  dyn_df=dyn_df.T
 
 
347
  dyn_df.columns = dyn_df.iloc[0]
348
  dyn_df=dyn_df.iloc[1:,:]
349