questions about ARIMA and auto ARIMA selection

[tarox]

hi,,

i used ARIMA model for forecasting data and i have questions about if i did use it correctly or not when use automatic arima selection.

1) I selected the series that i want (which is in level without take the first different or any change) and chose automatic arima selction , there is a choice about independent variable and there is c on it should i leave it ,,?

2) i left all what arimasel suggested (with independent c ) which are: AR(2),MA(2), SAR(12), SMA(6) ,,,SO on estimate equation i wrote the following :

d(y) c AR(2) MA(2) ....> ok ,,,,the result for correlogram -q statistics showed positive auto-correlation so i re estimate the previews equation as the following :

d(y) c AR(2) MA(2) SAR(12)

and the result showed perfect model, are my step correct?? :?

3) if i want to add x to the model and it is stationery at first difference should i on window arimasel on independent variable add d(x) ?

4)for estimation equation i wrote as suggested on automatic arima selection when added d(x) on independent the following:

d(y) c d(x) AR(2) MA(3) and if the result for correlogram -squared residuals showed positive auto-correlation should i re estimate the previews equation as the following:

d(y) c d(x) AR(2) MA(3) SAR(12) or i should correct the model by other ways ?

5) which series should i apply automatic arima section the series in level or the series in first different?

I would greatly appreciate any help that anyone can give :(