[Limdep Nlogit List] Switching Regression without the SWITCH Command

Alok R. Saboo arsaboo at gmail.com
Sat Nov 6 03:29:50 EST 2010


Hi,

I am taking another approach to the problem that I posted earlier and now
plan to run a Switching Regression with a Separator. I am using the code
provided on E23-64 (reproduced below).

SWITCH ;LHS = CAR11
;RH1 = ONE,MBA
;RH2 = ONE,MBA
;SEP=ORIGIN $

The problem, however, is that I want to model latent classes on the
intercept term for the two regimes, i.e., for the two RH2's in the last
statement, I want to be able to use latent classes. However, the ;LCM
command does not work with the SWITCH command.

I was wondering if I can replicate the SWITCH command using the basic
REGRESS commands, I should be able to use the LCM option. However, I am not
sure how can I replicate the SWITCH command using some of the basic
commands.

Would appreciate some help here....thanks!!

Thanks & Regards,

*Alok R. Saboo*
Doctoral Candidate in Marketing
Smeal College of Business,
Pennsylvania State University,
University Park, PA 16802.


More information about the Limdep mailing list