[Limdep Nlogit List] Models can not run
Thao Thai
Thao.T.Thai at monash.edu
Fri Jun 12 12:51:51 AEST 2020
Dear Nlogit users,
I have a DCE design with 6 attributes and the syntax of the MNL model is at
the bottom of this message.
I have run a number of models, of which I can not run some of them.
CL model Heteroscedastic MNL ML Generalised MNL
Attribute only AIC Can run Status=1 with maxit=200 Can run Can run
Attributes and a subset of 6 SDCs AIC Can run Status=1 with maxit=200 Can
run Status=1 with maxit=200
Attributes and more than 6 covariates and interaction terms Can run
Status=1
with maxit=200 Status=1 with maxit=200 Status=1 with maxit=200
Notes: MX and GMX model: 50 Halton draws
When I run the *nested logit model*, only one structure (;tree=
In(H,C,P,I,G)],Out(N)) can run (STATUS=0) while others do not converge
(STATUS=1) with the message "Fixed parameter ... is constrained to equal
the value or had a nonpositive st.error because of an earlier problem."
When I run the *latent class* with 2 classes, STATUS =1 with maxit =100.
When I tried the 3 class LC model, I got the message "Maximum of 100
iterations. Exit iterations with status=1
Error 1027: Models - estimated variance matrix of estimates is singular"
The syntax of MNL is below and it run fine
Nlogit
;lhs = cho, cset, alti
;choices = H, C, P, I, G, N
;model:
U(H) = rl_h1 * RL_H1 + rl_h2 * RL_H2
+ fl_ * FL_H
+ cr_h1 * CR_H1 + cr_2 * CR_H2
+ lo_h1 * LO_H1
+ sa_ * SA_H
/
U(C) = com
+ rl_c1 * RL_C1 + rl_c2 * RL_C2
+ fl_ * FL_C
+ cr_c1 * CR_C1 + cr_2 * CR_C2
+ lo_c1 * LO_C1 + lo_c2 * LO_C2
+ sa_ * SA_C/
U(P) = pri
+ rl_p1 * RL_P1
+ fl_ * FL_P
+ cr_p1 * CR_P1 + cr_2 * CR_P2
+ lo_p1 * LO_P1 + lo_p2 * LO_P2
+ sa_ * SA_P/
U(I) = ind
+ rl_i1 * RL_I1 + rl_i2 * RL_I2
+ fl_ * FL_I
+ cr_i1 * CR_I1
+ lo_i1 * LO_I1
+ sa_ * SA_I/
U(G) = gov
+ rl_g1 * RL_G1
+ fl_ * FL_G
+ cr_g1 * CR_G1
+ lo_g1 * LO_G1
+ sa_ * SA_G/
U(N) = non
+ rl_n1 * RL_N1
+ fl_ * FL_N
+ cr_n1 * CR_N1
+ lo_n1 * LO_N1 + lo_n2 * LO_N2
+ sa_ * SA_N
$
I am not sure what went wrong. I would be very grateful if you could
provide any insights on the problems that I am facing. Thank you so much in
advance.
Best regards,
T
More information about the Limdep
mailing list