import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_absolute_error


data = pd.read_csv('/datasets/energy_consumption.csv', index_col=[0], parse_dates=[0])
data.sort_index(inplace=True)
data = data.resample('1D').sum()

def make_features(data, max_lag, rolling_mean_size):
    data['year'] = data.index.year
    data['month'] = data.index.month
    data['day'] = data.index.day
    data['dayofweek'] = data.index.dayofweek

    for lag in range(1, max_lag + 1):
        data['lag_{}'.format(lag)] = data['PJME_MW'].shift(lag)

    data['rolling_mean'] = data['PJME_MW'].shift().rolling(rolling_mean_size).mean()


make_features(data, 5, 20)

train, test = train_test_split(data, shuffle=False, test_size=0.2)
train = train.dropna()

# < напишите код здесь >

# < обучите модель >
model = LinearRegression().fit(train.drop('PJME_MW', axis=1), train['PJME_MW'])

# < напишите код здесь >
pred_train = model.predict(train.drop('PJME_MW', axis=1))
pred_test = model.predict(test.drop('PJME_MW', axis=1))
print("MAE обучающей выборки:", mean_absolute_error(train['PJME_MW'], pred_train))
print("MAE тестовой выборки: ", mean_absolute_error(test['PJME_MW'], pred_test))