07-10-2012 05:23 PM
I'm new to SAS, and have a quick question. I would like to fit a Heckman selection correction model with a large number of parameters (3000+). I know that SAS works well on large datasets (I have 1.5 million observations), but I'm not sure how it does with large numbers of parameters. Can you help me with this question? Also, where would I have been able to look up the answer to this simple question on the SAS website? Thanks.
07-11-2012 08:09 AM
If you have SAS/ETS licensed, take a look at PROC QLIM. The online documentation should give examples (22.4 and 22.5), and the Details: QLIM Procedure section has a tab on selection methods. I would be very apprehensive about the 3000+ parameters, though.