03-21-2012 07:18 PM
I am currently working on a project where the effects of Easter have to be removed:
I am using Proc X12's regARIMA modelling and have a regression parameter of 14347.2733 as the Regression Parameter (for the pre-determined interval of 7 days)
I was wondering if anyone could help me understand how the regARIMA holiday components table is calculated?
For example in April 1987, the whole interval was in April and the table has given me a value of -4826.4 for March and 4826.4 for April and then these numbers are used to adjust the series prior to the seasonal adjusting I think!
Here is the code used:
title 'Removal of Easter Effect from March and April';
proc x12 data=textiles date=sasdate MdlInfoOut=mdlout1;
automdl maxdiff = (1,1);
forecast lead = 12;
output out=out1 a1 a7 b1;
Any help at all would be greatly appreciated.