topic Re: Proc surveyreg vs proc reg vs proc model etc in SAS Programming

Proc surveyreg vs proc reg vs proc model etc

somebody — Tue, 26 Nov 2019 07:23:28 GMT

I have a general question regarding procedures that run regressions. I work with financial data, especially asset pricing. I ran into many procedures that seem to perform the same task of running regressions, namely PROC SURVEYREG, PROC REG, PROC MODEL, PROC AUTOREG. Does anyone know, or know of, a summary or comparison between these PROCs? What are the main advantages/disadvantages of each one?

Re: Proc surveyreg vs proc reg vs proc model etc

PaigeMiller — Tue, 26 Nov 2019 11:51:23 GMT

Yes, there is certainly some overlap in the features of these procedures.

The SAS documentation contain descriptions of what each PROC does, for example from PROC AUTOREG documentation:

The AUTOREG procedure estimates and forecasts linear regression models for time series data when the errors are autocorrelated or heteroscedastic. The autoregressive error model is used to correct for autocorrelation, and the generalized autoregressive conditional heteroscedasticity (GARCH) model and its variants are used to model and correct for heteroscedasticity.

So you should start there in trying to learn what the procedures do.

Re: Proc surveyreg vs proc reg vs proc model etc

ballardw — Tue, 26 Nov 2019 16:00:29 GMT

https://documentation.sas.com/?cdcId=statcdc&cdcVersion=14.2&docsetId=statug&docsetTarget=statug_introreg_toc.htm&locale=en

The SAS/STAT documentation has a section on introduction to regression procedures.