[Limdep Nlogit List] RPM binary logit and probit

Stephen Beville stephenbeville at gmail.com
Fri May 23 14:42:21 EST 2008


Hi Pascal

This this the best I can do:


Firstly, I'm not sure why you have a ",R" command after your pts command?
Is this a typo?

Secondly, I would recommend renaming all your coefficient names eg. Q9_11
if you've used the exact same name specification for your variables.  For
instance if your variabe name is Q9_11 and you've named your coefficient the
same thing, Q9_11 . Don't do this.  Ive found that Nlogit can sometimes get
fowled up by this.

Best of luck,
SB

I


On 5/23/08, pascal bourgeat <pascalbourgeat at yahoo.com.au> wrote:
>
> As requested , I have pasted the code I have used for this analysis.  beats
> me really why the RPM command does work for probit but not for logit.  For
> logit, limdep just runs the standard model and then stops, the random
> coeeficient model is not estimated.
>
> LOGIT;Lhs=SATQ7_3;Rhs=ONE,Q9_1,Q9_10,Q9_11,Q9_12,Q9_13,Q9_14,Q9_15,Q9_16
>    ,Q9_17,Q9_2,Q9_3,Q9_4,Q9_5,Q9_6,Q9_7,Q9_8,Q9_9$
>
> PROBIT;Lhs=SATQ7_3;Rhs=ONE,Q9_1,Q9_10,Q9_11,Q9_12,Q9_13,Q9_14,Q9_15,Q9_16
> ,Q9_17,Q9_2,Q9_3,Q9_4,Q9_5,Q9_6,Q9_7,Q9_8,Q9_9; prob=prob1$
>
> LOGIT;Lhs=SATQ7_3;Rhs=ONE,Q9_1,Q9_10,Q9_11,Q9_12,Q9_13,Q9_14,Q9_15,Q9_16
>    ,Q9_17,Q9_2,Q9_3,Q9_4,Q9_5,Q9_6,Q9_7,Q9_8,Q9_9;
> RPM; fcn = one(N) , Q9_10(N), Q9_12 (N), Q9_13 (N), Q9_15 (N), Q9_4 (N),
> Q9_5 (N), Q9_8 (N) ;par; pts=25, R; halton $
>
> PROBIT;Lhs=SATQ7_3;Rhs=ONE,Q9_1,Q9_10,Q9_11,Q9_12,Q9_13,Q9_14,Q9_15,Q9_16
>    ,Q9_17,Q9_2,Q9_3,Q9_4,Q9_5,Q9_6,Q9_7,Q9_8,Q9_9; RPM; fcn = one(N) ,
> Q9_10(N), Q9_12 (N), Q9_13 (N), Q9_15 (N), Q9_4 (N), Q9_5 (N), Q9_8 (N) ;
> par; pts=25, R; halton;prob=prob2 $
>
>
> thanks for any feedback.
>
> pascal
>
>
>
> ---------------------------------
> Get the name you always wanted with the new y7mail email address.
> _______________________________________________
> Limdep site list
> Limdep at limdep.itls.usyd.edu.au
> http://limdep.itls.usyd.edu.au
>



-- 
Stephen Beville
PhD Candidate
Environment Society & Design Division
Lincoln University
(027) 338 8438
<*{:):):):><{



More information about the Limdep mailing list