Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: RE: How to deseasonalize with Stata

Subject   Re: st: RE: How to deseasonalize with Stata
Date   Mon, 29 Jul 2002 11:35:08 -0300

Thanks Nick for your reply.
I'm dealing with economic data, like gdp, investment, etc (quarterly data).
My understanding is that the most 'famous' method used by economists is the
X11ARIMA, originally developed by Statistics Canada in 1980 with later
changes and enhancements made in 1988. If Iīm not mistaken, this method is
already embedded in EViews and SAS, and I was wondering whether somebody
had written something similar to this for Stata, which is my favorite

> Iīve been trying to find a routine to deseasonalize data with Stata but I
> couldnīt find one.
> Can someone tell me if there exists a user written program
> to do this?

My understanding is that there are many ways to
"deseasonalise". Also, I don't pick up a hint of
what kind of data you are dealing with. Some
people do it with Fourier analysis,
and you don't need special Stata tools for that,
as it may boil down to a -regress- or -glm-.

I looked at what there is with

findit seasonal

which seemed to find only graphics routines
and some tests.

Some of the simpler methods count as linear filters
and so can be implemented via -egen, filter()-.
The -filter()- function is user-written
and downloadable via

ssc inst egenmore

Some of the comments at

How can I calculate moving averages for panel data?

-- even though it's ostensibly on a different question --
carry over to this question. In particular, if you have
-tsset- your data, then doing it yourself
may be much easier than you fear.

Alternatively, you might be able to do this via fitting an
appropriate model using -arima- or
-arch-, i.e. fit a model including seasonal


*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index