return dataZ
subj_list = [
'P103', 'P105', 'P107', 'P108', 'P110', 'P114', 'P116', 'P118', 'P120',
'P121'
]
for subj in subj_list:
print("************************")
print("Testing on {}".format(subj))
personalized_folder = '../data/wild/{}/personalized'.format(subj)
maybe_create_folder(personalized_folder)
df = pd.read_csv('../data/wild/{}/feature.csv'.format(subj))
exclusion = ['label', 'date_exp', 'start', 'end']
feature_names = [f for f in df.columns.values if not f in exclusion]
selected_columns = feature_names[:-5]
df = standardize_zscore(df, selected_columns)
# Leave one day out validation
day = []
day_index = []
precision = []
recall = []
fscore = []
precision_neg = []
return dataZ
subj_list = [
'P103', 'P105', 'P107', 'P108', 'P110', 'P114', 'P116', 'P118', 'P120',
'P121'
]
for subj in subj_list:
print("************************")
print("Testing on {}".format(subj))
generalized_folder = '../data/wild/{}/generalized'.format(subj)
maybe_create_folder(generalized_folder)
df = pd.read_csv('../data/wild/{}/feature.csv'.format(subj))
exclusion = ['label', 'date_exp', 'start', 'end']
feature_names = [f for f in df.columns.values if not f in exclusion]
selected_columns_normalization = feature_names[:-5]
concat_list = []
for outside in subj_list:
if outside == subj:
print("Excluding {}".format(outside))
continue
df_outside = pd.read_csv('../data/wild/{}/feature.csv'.format(outside))
concat_list.append(df_outside)
df_generalized = pd.concat(concat_list)
for subj in subj_list:
files = list_files_in_directory(
os.path.join('../data/wild/', subj, 'personalized'))
dur_df_list = []
# for file in files:
for day in range(14):
filepath = "../data/wild/{}/personalized/prediction_day_{}.csv".format(
subj, day)
if not os.path.isfile(filepath):
continue
maybe_create_folder(os.path.join('Meal_prediction', subj))
lprint(log_file, filepath)
raw_df = pd.read_csv(filepath)
# TIME ZONE CONTAINED AUTO CHECK CODE:
raw_df = df_to_datetime_tz_aware(raw_df, ['start', 'end'])
starttimes = raw_df['start'].values
endtimes = raw_df['end'].values
preds = raw_df['prediction'].values
earliest = raw_df['start'].min()
latest = raw_df['end'].max()
dt = pd.date_range(start=earliest, end=latest, freq='50ms')