import pandas as pd
import os
import re
file_dir1 = r'C:\Users\鸽子\Desktop\一版结果\电压等级电量预测结果\偏差率'
file_dir2 = r'C:\Users\鸽子\Desktop\一版结果\电压等级电量预测结果\月底3天预测结果'
file_dir3 = r'C:\Users\鸽子\Desktop\一版结果\行业电量预测结果\偏差'
import numpy as np
np.set_printoptions(threshold=np.inf)

# print(os.listdir(file_dir3))
# str1 = '丽水电压等级10kv以下月底偏差率:0.00229'
#
# print(re.split('电压等级|月底偏差率:',str1))
# with open(os.path.join(file_dir3,'9月底偏差率.txt'),'r',encoding='utf-8') as f:
#     lines = f.readlines()
#     list_city = []
#     list_industry = []
#     list_loss = []
#     for i in lines:
#         i = re.split(':|：|其中', i)
#         print(i)
#         list_city.append(i[0][:2])
#         list_industry.append(i[-2].replace(i[0][:2],''))
#         list_loss.append(i[-1][:-2])
# df_level = pd.DataFrame({'城市':list_city,'行业':list_industry,'偏差':list_loss})
# # df_level.to_csv(os.path.join(file_dir3,'9月底偏差率.csv'),encoding='gbk')
# print(df_level)
file_dir = r'C:\python-project\pytorch3\浙江行业电量\浙江所有地市133行业数据'
# print(os.listdir(file_dir))
dict1 = {}

for file in os.listdir(file_dir):

    df = pd.read_excel(os.path.join(file_dir,file),index_col=' stat_date ')

    col_list = df.drop(columns=[i for i in df.columns if (df[i] == 0).sum() / len(df) >= 0.5]).columns
    dict1[file[:2]] = col_list
    print(dict1)

    # print(len(df.drop(columns=[i for i in df.columns if (df[i] == 0).sum() / len(df) >= 0.5]).columns))

read_path = r'C:\Users\鸽子\Desktop\一版结果\行业电量预测结果\月底预测结果'
list1 = []
for i in os.listdir(read_path):
    print(i)
    data = pd.read_csv(os.path.join(read_path, i), sep='\t',header=None)
    data = data[data.columns[1:]]


    for j,step in enumerate(range(0, len(data), 4)):
        df = data.iloc[step+1:step + 4, :]
        df.columns = ['预测值', '实际值', '偏差率']
        try:
            df['行业'] = dict1[i[2:4]][j]
        except:
            pass
        df['城市'] = i[2:4]
        list1.append(df)
        print(df)
df = pd.concat(list1,ignore_index=True)
df.to_csv('各市行业电量预测结果.csv',encoding='gbk')
print(df)