how to set the order=(0,1,0), seasonal_order=(0,1,1,12) in the model

[tianke0711]

Hi, very nice tutorial, it is helpful to me . thanks! I am a new one to ARIMA model
I want to ask some questions about your tutorial.

1. Could you explain how to set the parameter value such as order=(0,1,0), seasonal_order=(0,1,1,12), could you tell me in detail, I found you make the data to be stationary , but you build the model by the df.riders data not used df.seasonal_first_difference data which have been stationary. Could you tell me how to get the parameter in detail.

mod = sm.tsa.statespace.SARIMAX(df.riders, trend='n', order=(0,1,0), seasonal_order=(1,1,1,12))

2. I found you used all date range data (df.riders) to fit the model, when you validating you model use the data range((start = 102, end= 114) which is a part of df.riders data. I mean why you didn't like the machine learning build the mode by the training data, and test the model by the test data. I mean the data should be split by two parts: fit mode data and test model data. why the SARIMAX use all data to fit model.
   Could you tell me in detail, please!

I hope you can help me to understand the above question, thanks!

[ahmed915]

Dear,
You get them by, Grid Search framework for Time Series Forecasting.
you have:
Grid search framework for Exponential smoothing, ARIMA model, and SARIMAX too.
feel free to contact me if you need to know more.

Good luck,

Ahmed