[Limdep Nlogit List] Multiple SP questions per respondents and Starting Values in RPL model

[Philip A. Viton]

On the starting values question: just use the ordinary "start = " 
specification. Note (1) that it is not possible to provide starting 
values for only a subset of the parameters : it's all or nothing; and 
(2) NLOGIT will typically re-arrange the independent variables from 
your lhs specification. I'm pretty sure --- someone correct me if I'm 
wrong --- that the re-arranged order is the one you want. This means, 
in particular, that the position of the start values will typically 
change over different random parameters specifications, since that 
leads to different orders of the indep vars.




At 06:27 PM 7/27/2007, you wrote:
>Dear All,
>
>
>
>I am trying to estimate a random parameter logit model for predicting
>mode choice of drivers. The data contains four stated preference
>questions corresponding to four trips for every traveler. I was
>wondering how to account for this, considering that we assume the
>distribution of random parameter across individuals and not across all
>trips.
>
>
>
>Also, if I wish to provide starting values for the random parameter (for
>both mean and std. deviation) how can it be done?
>
>
>
>I would greatly appreciate any response to these queries.
>
>
>
>Thanks,
>
>Sunil Patil
>
>PhD Candidate,
>
>Texas A & M University,
>
>College Station, Texas, USA
>
>
>
>
>
>_______________________________________________
>Limdep site list
>Limdep at limdep.itls.usyd.edu.au
>http://limdep.itls.usyd.edu.au


------------------------
Philip A. Viton
City Planning, Ohio State University
275 West Woodruff Avenue, Columbus OH 43210
viton.1 at osu.edu