'subsample':1,
'silent':1,
'nthread':27,
# 'scale_pos_weight':y_build.mean(),
# 'eval_metric':'auc',
'eval_metric':'logloss',
'objective':'binary:logistic',
'tree_method':'hist'}
print("""#==== print param ======""")
print('seed:', seed)
#==============================================================================
# prepare
#==============================================================================
train = pd.concat([utils.load_pred_None('trainT-0', W, True).sample(frac=.2),
utils.load_pred_None('trainT-1', W, True).sample(frac=.2),
utils.load_pred_None('trainT-2', W, True).sample(frac=.2)
], ignore_index=True, join='inner')
sub_train = train[['order_id', 'y']]
y_train = train['y']
X_train = train.drop('y', axis=1)
del train; gc.collect()
# drop id
col = [c for c in X_train.columns if '_id' in c] + ['is_train']
col.remove('user_id')
print('drop1',col)
X_train.drop(col, axis=1, inplace=True) # keep user_id
'silent': 1,
'nthread': 28,
'eval_metric': 'logloss',
'objective': 'binary:logistic',
'tree_method': 'hist'
}
print("""#==== print param ======""")
print('DATE:', DATE)
print('seed:', seed)
#==============================================================================
# prepare
#==============================================================================
train = pd.concat([
utils.load_pred_None('trainT-0', 3),
utils.load_pred_None('trainT-1', 3),
utils.load_pred_None('trainT-2', 3)
],
ignore_index=True)
sub_train = train[['order_id', 'y']]
y_train = train['y']
X_train = train.drop('y', axis=1)
del train
gc.collect()
# drop id
col = [c for c in X_train.columns if '_id' in c] + ['is_train']
col.remove('user_id')
print('drop1', col)
'subsample':0.75,
'silent':1,
'nthread':28,
'eval_metric':'logloss',
'objective':'binary:logistic',
'tree_method':'hist'
}
print("""#==== print param ======""")
print('DATE:', DATE)
print('seed:', seed)
#==============================================================================
# prepare
#==============================================================================
train = pd.concat([utils.load_pred_None('trainT-0', 3),
utils.load_pred_None('trainT-1', 3),
utils.load_pred_None('trainT-2', 3)
], ignore_index=True)
sub_train = train[['order_id', 'y']]
y_train = train['y']
X_train = train.drop('y', axis=1)
del train; gc.collect()
# drop id
col = [c for c in X_train.columns if '_id' in c] + ['is_train']
col.remove('user_id')
print('drop1',col)
X_train.drop(col, axis=1, inplace=True) # keep user_id
