topic getting started with SEVERITY in SAS Forecasting and Econometrics

getting started with SEVERITY

rs_poetic — Tue, 24 Apr 2012 15:17:54 GMT

Hello,

Proc Severity, an experimental proc in ETS, sounds like the answer to my prayers, but I am stuck at the outset with what seems like a very simple error message. I am trying to run the example given in the documentation, simply:

proc severity data=spilldat;

model spill / crit = aicc;

run;

where spill is a variable in the data. Although the proc begins to run, the immediate error message in the log is:

180

errror 180-322: statement is not valid or it is used out of order

Please advise. (I have ETS as part of 9.3 TS level 1M1 on Windows 7.)

Re: getting started with SEVERITY

PGStats — Tue, 24 Apr 2012 15:37:52 GMT

This is a typical syntax error message. Looking at the doc, there is no MODEL statement for that proc and the CRIT option belongs in the PROC statement. - PG

Re: getting started with SEVERITY

rs_poetic — Tue, 24 Apr 2012 16:18:58 GMT

Thank you. The MODEL statement does look peculiar but perhaps that is due to the context being to fit a predefined univariate probability distribution. However, the example in the documentation on page 1 of "A simple Example of Fitting Predefined Distributions" is clear that "the MODEL statement specifies the variable to be modeled along with the model selection criterion" and I copied the example there exactly (I believe). Can you provide an illustrative little program for proc severity that does not include a MODEL statement?

Re: getting started with SEVERITY

PGStats — Tue, 24 Apr 2012 17:29:46 GMT

That example is from SAS version 9.22. It seems that the syntax was changed in version 9.3 (it IS experimental!)

http://support.sas.com/documentation/cdl/en/etsug/63939/HTML/default/viewer.htm#etsug_severity_sect008.htm

Re: getting started with SEVERITY

rs_poetic — Tue, 24 Apr 2012 17:33:32 GMT

Thanks!!!