Statsmodels

[garikhgh0]

hi, I have q question on how to define p,q and q values for Arima Models.

fig = plt.figure(figsize=(16,6))
model = ARIMA(ts_log, order=(17, 1, 5))  
results_AR = model.fit(disp=-1)  
plt.plot(ts_log_diff)
plt.plot(results_AR.fittedvalues, color='red')
# plt.title('RSS: %.4f'% sum((results_AR.fittedvalues-ts_log_diff)**2))

as the pridicting result is very disappointing
the orange line it the prediction which is not what I seek, I have tried with many different values of p,d and q, it does not give the result.
thanks in advance