import xgboost as xgb
import pandas as pd
import os
from sklearn.metrics import r2_score
from sklearn.model_selection import train_test_split
pd.set_option('display.width',None)
def season(x):
    if str(x)[5:7] in ('01', '02'):
        return 0
    elif str(x)[5:7] in ('03', '04', '05', '06', '09', '10', '11', '12'):
        return 1
    else:
        return 2

def normal(nd):
    high = nd.describe()['75%'] + 1.5*(nd.describe()['75%']-nd.describe()['25%'])
    low = nd.describe()['25%'] - 1.5*(nd.describe()['75%']-nd.describe()['25%'])
    return nd[(nd<high)&(nd>low)]

parent_dir = os.path.abspath(os.path.join(os.getcwd(),os.pardir))
data = pd.read_excel(os.path.join(parent_dir,'入模数据/衢州.xlsx'))
data['dtdate'] = pd.to_datetime(data['dtdate'],format='%Y-%m-%d')
data['year'] = data['dtdate'].dt.year

data.set_index('dtdate',inplace=True)
data = data.loc[normal(data['售电量']).index]

# list2 = []
# list0 = []
# list1 = []
# for i in ('01','02','03','04','05','06','07','08','09','10','11','12'):
#     month_index = data.index.strftime('%Y-%m-%d').str[5:7] == f'{i}'
#     if data.loc[month_index]['售电量'].mean() >= data['售电量'].describe()['75%']:
#         list2.append(i)
#     elif data.loc[month_index]['售电量'].mean() <= data['售电量'].describe()['25%']:
#         list0.append(i)
#     else:
#         list1.append(i)
#
# print(list0,list1,list2)
data['season'] = data.index.map(season)

df_eval = data.loc['2023-12']
df_train = data.iloc[-180:]
print(df_train)

df_train = df_train[['tem_max','tem_min','holiday','24ST','售电量','season','year']]

X = df_train[['tem_max','tem_min','24ST','holiday','season','year']]
X_eval = df_eval[['tem_max','tem_min','24ST','holiday','season','year']]
y = df_train['售电量']

x_train,x_test,y_train,y_test = train_test_split(X,y,test_size=0.2,random_state=42)
model = xgb.XGBRegressor(max_depth=6, learning_rate=0.1, n_estimators=150)
model.fit(x_train,y_train)
y_pred = model.predict(x_test)
result_test = pd.DataFrame({'test':y_test,'pred':y_pred},index=y_test.index)

# 指标打印
eval_pred = model.predict(X_eval)
result_eval = pd.DataFrame({'eval':df_eval['售电量'],'pred':eval_pred})

goal = (result_eval['eval'][-3:].sum()-result_eval['pred'][-3:].sum())/result_eval['eval'].sum()
print(goal)

goal2 = (result_eval['eval'][-23:].sum()-result_eval['pred'][-23:].sum())/result_eval['eval'].sum()
print(goal2)
print(result_eval)

import numpy as np
X_eval = np.array([
    [15.7, 4.5, 0, 0, 1, 2023],
    [16.0, 4.8, 0, 0, 1, 2023],
    [17.0, 7.4, 0, 3, 1, 2023],
    [12.9, 4.1, 0, 3, 1, 2023]
])
print(model.predict(X_eval))
result = model.predict(X_eval)
result = pd.DataFrame(result, index=['2023-12-28', '2023-12-29', '2023-12-30', '2023-12-31'])
result = pd.concat((result_eval['eval'],result))
result.index = result.index.map(lambda x:str(x)[:10])
result.columns = ['预测值']
print(result)

with pd.ExcelWriter(r'C:\Users\鸽子\Desktop\市区日电量预测_1229.xlsx', mode='a', if_sheet_exists='replace',
                    engine='openpyxl') as writer:
    result.to_excel(writer, sheet_name='衢州')

# import torch
# from torch import nn
# from torch.utils.data import TensorDataset,DataLoader
#
#
# net = nn.Sequential(
#     nn.Linear(6,32),
#     nn.ReLU(),
#     nn.Linear(32,64),
#     nn.ReLU(),
#     nn.Linear(64,64),
#     nn.ReLU(),
#     nn.Linear(64,1)
# )
# opt = torch.optim.Adam(net.parameters(),lr=0.00005)
# loss_fn = nn.MSELoss()
#
# epochs = 200
#
# x_train = torch.from_numpy(x_train.values).type(torch.float32)
# x_train = (x_train - x_train.mean())/x_train.std()
#
# y_train = torch.from_numpy(y_train.values).type(torch.float32)
# std1 = y_train.std()
# mean1 = y_train.mean()
# y_train = (y_train - mean1)/std1
#
# X_eval = torch.from_numpy(X_eval.values).type(torch.float32)
# X_eval = (X_eval - X_eval.mean())/X_eval.std()
#
# y_eval= torch.from_numpy(df_eval['售电量'].values).type(torch.float32)
#
#
# train_ds = TensorDataset(x_train,y_train)
# train_dl = DataLoader(train_ds,shuffle=True,batch_size=64)
#
#
# for i in range(epochs):
#     for x,y in train_dl:
#         y_pred = net(x)
#         loss = loss_fn(y_pred,y)
#
#         opt.zero_grad()
#         loss.backward()
#         opt.step()
#         print(round(loss.item(),2))
#
# predict = (net(X_eval) * std1 + mean1).detach().numpy()
# print(y_train.std(),y_train.mean())
# print(net(X_eval))
# print(predict)
# print((y_eval.detach().numpy().sum() - predict.sum())/ y_eval.detach().numpy().sum())