'DBD': ['max', 'mean', 'sum'],
'AMT_INSTALMENT_sub_AMT_PAYMENT': ['max', 'mean', 'sum', 'var'],
'AMT_INSTALMENT_div_AMT_PAYMENT': ['max', 'mean', 'sum', 'var'],
'AMT_INSTALMENT': ['max', 'mean', 'sum'],
'AMT_PAYMENT': ['min', 'max', 'mean', 'sum'],
'DAYS_ENTRY_PAYMENT': ['max', 'mean', 'sum']
}
via = df.groupby('SK_ID_PREV').agg(aggs).reset_index(drop=True)
via.columns = [
'_'.join(f) if f[0] != 'SK_ID_CURR' else f[0] for f in via.columns
]
self.df = via.groupby('SK_ID_CURR').mean()
if __name__ == '__main__':
args = get_arguments('POS CASH')
with timer('load dataset'):
train = pd.read_feather(TRAIN)[['SK_ID_CURR']]
test = pd.read_feather(TEST)[['SK_ID_CURR']]
inst = pd.read_feather(INST)
with timer('preprocessing'):
# inst.drop(['SK_ID_PREV'], axis=1, inplace=True)
inst = inst.sort_values(['SK_ID_CURR',
'DAYS_INSTALMENT']).reset_index(drop=True)
inst.loc[:, inst.columns.str.startswith('AMT_')] = np.log1p(
inst.filter(regex='^(AMT_)'))
inst.loc[:, inst.columns.str.startswith('DAYS_')] = inst.filter(
regex='^DAYS_').replace({365243: np.nan})
with timer('create dataset'):
import os
import sys
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
from feat import get_arguments
FILES = ['main.py', 'bureau.py', 'credit_card_balance.py', 'installments_payment.py', 'pos_cash_balance.py',
'previous_application.py']
if __name__ == '__main__':
args = get_arguments('all')
for file in FILES:
cmd = ['python', '-u', file]
if args.force:
cmd += ['-f']
os.system(' '.join(cmd))