y_test_ids_df,
on=id_cols)
test_labs_df = pd.merge(collapsed_labs_df, y_test_ids_df, on=id_cols)
test_mews_df = pd.merge(mews_df, y_test_ids_df, on=id_cols)
# merge them
test_collapsed_features_df = pd.merge(test_vitals_df,
test_labs_df,
on=id_cols,
how='inner')
test_features_df = pd.merge(test_collapsed_features_df,
demographics_df,
on=id_cols)
if p == 0:
test_features_dict = merge_data_dicts([
collapsed_vitals_data_dict, collapsed_labs_data_dict,
demographics_data_dict
])
test_outcomes_df = pd.merge(test_features_df[id_cols],
outcomes_df,
on=id_cols,
how='inner')
# # get performance metrics
feature_cols = parse_feature_cols(test_features_dict['schema'])
mews_score_col = parse_feature_cols(mews_data_dict['schema'])
x_test = test_features_df[feature_cols].values
y_test = test_outcomes_df[outcome_col].values
mews_test = test_mews_df[mews_score_col].values
# bootstrap test set inds without replacement
chosen_stay_ids_df = chosen_stay_ids_df.drop_duplicates(subset=id_cols).reset_index(drop=True)
# for each patient get their vitals, labs, demographics
labs_df, labs_data_dict, vitals_df, vitals_data_dict, \
demographics_df, demographics_data_dict, outcomes_df, outcomes_data_dict = get_preprocessed_data(args.preproc_data_dir)
vitals = parse_feature_cols(vitals_data_dict)
labs = parse_feature_cols(labs_data_dict)
chosen_stay_labs_df = pd.merge(labs_df, chosen_stay_ids_df, on=id_cols, how='inner')
chosen_stay_vitals_df = pd.merge(vitals_df, chosen_stay_ids_df, on=id_cols, how='inner')
chosen_stay_highfreq_df = pd.merge(chosen_stay_labs_df, chosen_stay_vitals_df, on = id_cols + ['hours_since_admission',
'timestamp'], how='outer')
highfreq_features_dict = merge_data_dicts([labs_data_dict, vitals_data_dict])
highfreq_features_dict['fields'] = highfreq_features_dict['schema']['fields']
# choose a subject
chosen_short_stay_subj_list = ['14343967', '18115638', '18826316', '17245153',
'17557700', '11212084', '11163358', '17684794',
'12751842', '11528888', '1379931', '17745211',
'12862019', '14201044', '14917356', '17682462',
'1339889', '17995864', '15787542', '13007083',
'18239690', '11692208', '19352552', '19438165',
'14858518']
chosen_long_stay_subj_list = ['12702290', '19160387', '19806342', '19222313', '17863017']
chosen_stay_subj_list = chosen_short_stay_subj_list + chosen_long_stay_subj_list
# chosen_stay_subj_list = chosen_long_stay_subj_list[0:2]
for idx in chosen_stay_subj_list:
else:
test_collapsed_features_df = pd.merge(test_vitals_df,
test_labs_df,
on=id_cols,
how='inner')
data_dicts_list = [
collapsed_vitals_data_dict, collapsed_labs_data_dict,
demographics_data_dict
]
test_features_df = pd.merge(test_collapsed_features_df,
demographics_df,
on=id_cols)
if p == 0:
test_features_dict = merge_data_dicts(data_dicts_list)
test_outcomes_df = pd.merge(test_features_df[id_cols],
outcomes_df,
on=id_cols,
how='inner')
# # get performance metrics
feature_cols = parse_feature_cols(test_features_dict['schema'])
mews_score_col = parse_feature_cols(mews_data_dict['schema'])
x_test = test_features_df[feature_cols].values
y_test = test_outcomes_df[outcome_col].values
# load the scaler
scaler = pickle.load(
open(os.path.join(clf_models_dir, 'scaler.pkl'), 'rb'))
