def function_partial_autocorrelation(ts_data, window_size):
ts_log_diff = differencing_log(ts_data, window_size, False, False)
lag_pacf = pacf(ts_log_diff, nlags=20, method="ols")
pyplot.plot(lag_pacf)
pyplot.axhline(y=0, linestyle="--", color="gray")
pyplot.axhline(y=-1.96 / np.sqrt(len(ts_log_diff)),
linestyle="--",
color="gray")
pyplot.axhline(y=1.96 / np.sqrt(len(ts_log_diff)),
linestyle="--",
color="gray")
pyplot.title("Partial Autocorrelation Function")
def function_autocorrelation(ts_data, window_size):
ts_log_diff = differencing_log(ts_data, window_size, False, False)
lag_acf = acf(ts_log_diff, nlags=20)
pyplot.plot(lag_acf)
pyplot.axhline(y=0, linestyle="--", color="gray")
pyplot.axhline(y=-1.96 / np.sqrt(len(ts_log_diff)),
linestyle="--",
color="gray")
pyplot.axhline(y=1.96 / np.sqrt(len(ts_log_diff)),
linestyle="--",
color="gray")
pyplot.title('Autocorrelation Function')
def model_combined(ts_data, window_size, should_plot=True):
ts_log = np.log(ts_data)
ts_log_diff = differencing_log(ts_data, window_size, False, False)
model = ARIMA(ts_log, order=(1, 1, 1))
results_ARIMA = model.fit(disp=0)
if should_plot:
# root_mean_squared = sum((results_MA.fittedvalues - ts_log_diff) ** 2)
pyplot.plot(ts_log_diff, color="blue", label="ts log difference")
pyplot.plot(results_ARIMA.fittedvalues,
color="red",
label="results ARIMA fitted")
pyplot.legend(loc="best")
pyplot.title("Combined Model AR-I-MA")
# pyplot.title("RSS: {}".format(root_mean_squared))
return results_ARIMA