[R] Mixed conditional logit model

[Sven Müller]

dear all,

i wonder whether it is possible to estimate a mixed (random parameters) 
logit model in R. my dataset only includes conditional explanatory (RHS) 
variables. i've already searched the R-help archives and found slightly 
comparable questions but no satisfying answers. an old fashoined 
conditional logit does not work due to the violation of the iia 
property.  a short description of my dataset and task:
i'm trying to model the school-choice behaviour of students ('to which 
school a student attends'). since this model should be used for 
prediction it only handles conditional variables, such as profiles 
offered (sports, math, music etc) and the distance between the students 
home and the schools within the choice set. it is very important to keep 
the conditional structure of the model, because there will be new 
schools in the future, for which i've to predict the choice 
probabilities. thus, the only way to describe the attributes of these 
future schools is by distance and profile. i have to avoid choice 
specific parameters (like multinomial logit).
i'd appreciate any adive on packages, codes and examples.

thank you very much in advance ...

sven

-- 
Sven Mueller, M.A.

Dresden University of Technology
Institute for Economics and Traffic
Chair of Business Administration, Transport and Logistics

Andreas Schubert Str. 23,  01062 Dresden

Room 510

Phone: +49 351 463 36797
Fax: +49 351 463 37758

sven.mueller at tu-dresden.de
http://www.vbl-dresden.de