Forecasting using SAS Forecast Server, SAS/ETS, and more

FORECAST Procedure, Winters method smoothing parameters

Reply
New Contributor
Posts: 3

FORECAST Procedure, Winters method smoothing parameters

When you choose Winters method in FORECAST procedure you can select smoothing parameters. When you select them, by default SaS sets them to (1-0.8)^(1/trend). You can also define your own.

I wonder what are the values of smoothing parameters when you don't specify them explicitly. Any ideas where I can find it?

PROC Star
Posts: 1,167

Re: FORECAST Procedure, Winters method smoothing parameters

Posted in reply to KarolinaP

I suggest you post this in one of the more statistics-heavy areas, I think your chances of a reply will be better.

Tom

Community Manager
Posts: 2,953

Re: FORECAST Procedure, Winters method smoothing parameters

Posted in reply to KarolinaP

I moved this to where the experts will see it to respond.

Chris

New Contributor
Posts: 3

Re: FORECAST Procedure, Winters method smoothing parameters

Posted in reply to KarolinaP

Great, thanks for providing proper context!

Artur

SAS Employee
Posts: 17

Re: FORECAST Procedure, Winters method smoothing parameters

Posted in reply to KarolinaP

I have to admit I'm not 100 percent sure what you are asking. A more detailed description of what PROC FORECAST does can be found here: http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/etsug_forecast_details03.htm

However, if you are looking for a procedure that can optimize the smoothing weights for you based on data I would encourage you to look into PROC ESM which is another procedure in the SAS/ETS package - http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/etsug_esm_overview.htm

Thanks,

Snurre

New Contributor
Posts: 3

Re: FORECAST Procedure, Winters method smoothing parameters

Posted in reply to SnurreJensen

Thanks Snurre, I have seen both docs.

What I haven't find is whether FORECAST use ESM Procedure or other procedure to obtain smoothing parameters, when you don't specify them.

Artur

SAS Employee
Posts: 416

Re: FORECAST Procedure, Winters method smoothing parameters

Posted in reply to KarolinaP

Hello -

See section on "Smoothing Weights" here: http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/etsug_forecast_details03.htm when using the FORECAST procedure.

For the ESM procedure all the parameters (smoothing weights) associated with the exponential smoothing model used to forecast the time series are optimized based on the data, with the default parameter restrictions imposed. Details can be found here:  http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/viewer.htm#etsug_tffordet_sect0...

and here:

http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/viewer.htm#etsug_tffordet_sect0...

Thanks,

Udo

SAS Employee
Posts: 416

Re: FORECAST Procedure, Winters method smoothing parameters

Posted in reply to SnurreJensen

Hello -

If I understand this question correctly you may want to check out the OUTEST data set - see:

http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/viewer.htm#etsug_forecast_detai...

For WINTERS have a look at WEIGHT1 | WEIGHT2 | WEIGHT3. The observations contain the weights used for smoothing the WINTERS or ADDWINTERS method parameters. _TYPE_=WEIGHT1 is the weight used to smooth the CONSTANT parameter. _TYPE_=WEIGHT2 is the weight used to smooth the LINEAR and QUAD parameters. _TYPE_=WEIGHT3 is the weight used to smooth the seasonal parameters. These observations are output only for the WINTERS and ADDWINTERS methods.

As suggested you may want to check out PROC ESM instead of PROC FORECAST.

Thanks,

Udo

Ask a Question
Discussion stats
  • 7 replies
  • 1222 views
  • 0 likes
  • 5 in conversation