import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from IPython.display import display
from sklearn.linear_model import ElasticNetCV

# read the data
data = pd.read_csv('biggestmerchang.csv', parse_dates=['day'])
t = data['day']
y = data['activation']
test_size=0.2


# create features in new dataframe
X = pd.DataFrame()
X['daysbef0'] = y.shift(0)
X['daysbef5'] = y.shift(5)X['lag 10'] = y.shift(10)
X['daysbef10'] = y.shift(11)
X['t'] = (t - t.min())/(t.max() - t.min())
X['rollingmean3'] = y.rolling(3).mean()
X['rollingmean6'] = y.rolling(6).mean()
X['rollingmean12'] = y.rolling(12).mean()

train = X[:-test_size]
test  = X[-test_size:]

estimator = ElasticNetCV(n_alphas=100, l1_ratio=0.9, cv=5, random_state=123)
estimator.fit(X.loc[train, :], y[train])
test_score = estimator.score(X.loc[test, :], y[test])