|
|
|
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([
|
|
|
|
[15.0, 3.0, 0, 0, 1, 2023],
|
|
|
|
[15.7, 5.1, 0, 0, 1, 2023],
|
|
|
|
[15.2, 5.6, 0, 0, 1, 2023],
|
|
|
|
[17.9, 6.8, 0, 3, 1, 2023],
|
|
|
|
[12.0, 2.9, 0, 3, 1, 2023]
|
|
|
|
])
|
|
|
|
print(model.predict(X_eval))
|
|
|
|
result = model.predict(X_eval)
|
|
|
|
result = pd.DataFrame(result, index=['2023-12-27', '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_1228.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())
|
|
|
|
|
|
|
|
|
|
|
|
|