import matplotlib.pyplot as plt
import pylab
import matplotlib.mlab as mlab
import statsmodels.api as sm
from statsmodels.tsa.arima_model import ARIMA
import pandas as pd
import numpy as np
import scipy.stats as scipy_stats
import itertools
import warnings


def parser(x):
    return pd.datetime.strptime(x, '%Y-%m-%d')

p = d = q = range(0, 5)
pdq = list(itertools.product(p, d, q))

data = pd.read_csv('BCHAIN.csv', header=0, parse_dates=[0], index_col=0, squeeze=True, date_parser=parser)# import file dari csv
data = np.float64(data)

param = (1,1,1)
model = ARIMA(data, order=param)
result = model.fit()
print(result.aic)
print(result.summary())

param = (4,1,3)
model = ARIMA(data, order=param)
result = model.fit()
print(result.aic)
print(result.summary())

residuals = pd.DataFrame(result.resid)
residuals.plot(kind='kde')
print(residuals.describe())
plt.show()

forecast = result.forecast(steps=10, exog=None, alpha=0.05)
print(forecast[0])