#-*- coding:utf-8 -*-
__author__ = 'Oct. Chen'
"""
"""
from data import DataSet
from gbdt import GBDT
import numpy as np
if __name__ == '__main__':
data_file = './train.csv'
test_file = './test.csv'
dataset = DataSet(data_file,
test_file,
index_col='Id',
target_col='SalePrice')
#dataset = DataSet(data_file, index_col = 'ID', target_col = 'target')
dataset.drop_feature_by_missed_rate()
dataset.fix_na()
dataset.feature_discrete(skip_cols=['Id', 'SalePrice'])
# dataset.feature_discrete(skip_cols = ['ID', 'target'])
dataset.hold_feat_by_corr(0.11)
#asdsadad
dataset.target_trasfer()
gbdt = GBDT(300, 0.1, 0.3, 5, learn_rate=0.05)
test_output = gbdt.fit(dataset)
test_output = np.exp(np.array(test_output, dtype=np.float32)) - 1.
print(test_output)
