Hi!
I'm trying to run a logistic regression and implement LASSO variable selection using PROC HPGENSELECT. However, SAS did not produce confidence limits for the regression estimates even though I have requested it. Any idea what I did wrong? Thank you.
Panji
proc hpgenselect data=mydata;
class a b c;
model y(event='1')=x1 x2 x3 x4 x5/dist=binary link=logit cl;
selection method=lasso(choose=aicc);
run;
You should see a NOTE in the LOG alerting you to the fact that the CL option is not available when you use Lasso selection.
NOTE: The CL option is not available for the LASSO method.
Keep in mind that Lasso selection does not actually produce standard errors for the parameter estimates either, so it will not give you confidence limits.
What does your LOG show when you run the code, as in all of the messages or notes?
Did you intend to include those Class variables in the model?
You should see a NOTE in the LOG alerting you to the fact that the CL option is not available when you use Lasso selection.
NOTE: The CL option is not available for the LASSO method.
Keep in mind that Lasso selection does not actually produce standard errors for the parameter estimates either, so it will not give you confidence limits.
Available on demand!
Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.
ANOVA, or Analysis Of Variance, is used to compare the averages or means of two or more populations to better understand how they differ. Watch this tutorial for more.
Find more tutorials on the SAS Users YouTube channel.