[Limdep Nlogit List] random parameter bivariate ordered probit model error in NLOGIT 4.0...

Behram Wali walibehram at yahoo.com
Thu Feb 25 11:57:00 AEDT 2016


Dear all,
I am trying to estimate a random-parameter (rp) bivariate ordered probit model to investigate route choice decisions 
of two-individuals within same household... Can NLOGIT 4.0 estimate such a model? (as i could not find any material in manual)
I am currently trying this code but probably am not writing the code correctly.The software estimates the individual rp ord-probit models and store results in matrices, however i am unable to estimate 
the rp bivariate ord-probit model.

READ;Nobs=298;Nvar=78;File="C:\Users\Desktop\jj300.csv"$
create;if(x75=1)xx75=0$
create;if(x75=2)xx75=1$
create;if(x75=3)xx75=2$
create;if(x76=1)xx76=0$
create;if(x76=2)xx76=1$
create;if(x76=3)xx76=2$

oprobit;lhs=xx75;rhs=one,x32;RPM
;Pts=200;halton
;Fcn=x32(n)$
matrix;b1=b;mu1=mu$
oprobit;lhs=xx76;rhs=one,x32;RPM
;Pts=200;halton
;Fcn=x32(n)$
matrix;b2=b;mu2=mu$
oprobit;lhs=xx75,xx76
;rh1=one,x32
;rh2=one,x32
;start=b1,mu1,b2,mu2,0;Marginal effects$

"Error   715: The wrong number of starting values was found."
 Can anyone please guide me on this?
Thanks a lot please.

-Behram


More information about the Limdep mailing list