[Limdep Nlogit List] RP SP data and ECM

Brett Smith brett.smith at uwa.edu.au
Fri Sep 17 10:35:34 EST 2010


Hi group

I wonder if anyone can help me with using error components logit to combine RP-SP data. My initial thought is below, but I'm not too sure I am doing the right thing. Both st devs of the error components are very close to zero with large St errors +E10.




ECLOGIT
      ;lhs=CHOICE,CSET2,MODEalt
     ;choices=car,train,bus, rcar, rtrain, rbus
      ;ecm=(rcar,rtrain,rbus),(car,bus,train)
      ;halton
      ;pts=45
      ;pds=5
      ;model:
    U(car)=              V(car) /
    U(train)=   sAT   V(train) /
    U(bus)=     sAB   V(bus)  /
    U(rcar)=             V(rcar)/
    U(rtrain)=   rAT   + V(rtrain)/
    U(rbus)=     rAB   + V(rbus)$

Should I use Nlogit with random parameters on 1's?

Regards,


Brett Smith
UWA Business School
The University of Western Australia M261
35 Stirling Highway, Crawley
Perth, WA 6009
Australia
http://www.biz.uwa.edu.au/home/

Ph:  +61 +8 6488 3979
Fax:  +61 +8 6488 1055
Email: Brett.Smith at uwa.edu.au



More information about the Limdep mailing list