[SciPy-User] Seasonal adjustment in scipy/python?

josef.pktd@gmai... josef.pktd@gmai...
Thu Jan 14 13:59:57 CST 2010


On Thu, Jan 14, 2010 at 2:44 PM,  <totalbull@mac.com> wrote:
> Hello,
>
> I am looking to seasonally adjust some data series in Python - specifically
> economics in emerging markets. As you can see on the charts
> (www.emconfidential.com) there is a lot of seasonality to monthly data
> series. Example 1 (retail sales) is obvious. Example 2, CPI,  is somewhat
> less so, but there is still some seasonality here with price falls around
> January and fairly high prices in December.
>
> How would I go about seasonally adjusting this data using Python and Scipy?
> Any canned functions?

I haven't seen any canned functions, the simplest would be to use
annual differences or estimate the monthly base level with regression
on month dummy variables and take the residual.
>From the graph, it doesn't look like assuming a functional form for
the monthly base level (seasonal trend) would be useful.

There should be more sophisticated ways for filtering but not canned,
and I don't think X11 is available in python. It would also depend on
how long your timeseries is and what you want to do with it

Josef

>
> Tom
>
>
>
>
>
> _______________________________________________
> SciPy-User mailing list
> SciPy-User@scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-user
>
>


More information about the SciPy-User mailing list