[Limdep Nlogit List] How to fix certain coefficients at zero for Latent Class Model?

[Yuanchang Xie]

Dr. Greene,

Thank you for your help.  I tried Rst in my code. However, it did not
work out. The reason might be that I used NLOGIT command in the latent
class model. To confirm my guess, I did a simple experiment.

All explanatory variables I used are alternative specific variables. I
fitted two MNL models using MLOGIT and NLOGIT, respectively.  When the
MLOGIT command is used, the Rst can work properly and coefficients can
be prespecified. However, if the NLOGIT command is used, Rst
specifications have no effect on the coefficients, and no error
messages are given. Could you please run a simple test using the
latent class model example in the manual on Page N9-4?

Also, it sees that MLOGIT automatically chooses the first alternative
(Y=0) as the base alternative, and NLOGIT automatically uses the last
alternative as the base alternative. Is there a command that can set
the base alternative manually? Thank you so much.

Yuanchang