[Limdep Nlogit List] Error message in running Stated Choice Model

[Puri,Mahi]

Hello

I am using NLOGIT (version 6) and trying to run Mixed (Random Parameter) Logit Model for stated choice experiment data. However, the model is giving an error and we are unable to proceed. Please advice.

The code, description of data and error message are provided below:

  1.
The data is in the form of panel data (3 repetitions) where each respondent is given 3 choices (alt 1, alt 2, none). So, 600 respondents*3 choices*3 repetitions = 5400 rows of data
  2.
The attributes (land, year, payment) vary over choices.
  3.
Covariates/characteristics (padist, fcov, herbcon, inputs, benefits, respage, respedu, hhmem, #caste, history, landsize, aginc, crops, otherinc, sources) vary across respondents.
  4.
"land" and "year" are random parameters

|-> NLOGIT; Lhs = choiced

    ; Choices = alt1, alt2, none

           ; model:

    U(alt1) = b_land * land + b_year * year + b_payment * payment/

    U(alt2) = b_land * land + b_year * year + b_payment * payment/

    U(none) = asc + b_land * land + b_year * year + b_payment * payment

           ; RPL = padist, fcov, herbcon, inputs, benefits, respage, respedu, hhmem, #caste, history, landsize, aginc, crops, otherinc, sources

           ; Pts = 500; Halton; Pds = 3

           ; Fcn = b_land(n), b_year(n)$


-----------------------------------------------------------------------------

Start values obtained using MNL model

Dependent variable               Choice

Log likelihood function     -1588.38429

Estimation based on N =   1773, K =   4

Inf.Cr.AIC  =   3184.8 AIC/N =    1.796

---------------------------------------

            Log likelihood R-sqrd R2Adj

Constants only  -1913.0514  .1697 .1602

Note: R-sqrd = 1 - logL/Logl(constants)

Warning:  Model does not contain a full

set of ASCs. R-sqrd is problematic. Use

model setup with ;RHS=one to get LogL0.

---------------------------------------

Response data are given as ind. choices

Number of obs.=  1800, skipped   27 obs

--------+--------------------------------------------------------------------

        |                  Standard            Prob.      95% Confidence

CHOICED|  Coefficient       Error       z    |z|>Z*         Interval

--------+--------------------------------------------------------------------

  B_LAND|    -.04070***      .00181   -22.51  .0000     -.04425   -.03716

  B_YEAR|    -.02464**       .01144    -2.15  .0313     -.04707   -.00221

B_PAYMEN|     .01460***      .00225     6.50  .0000      .01020    .01900

     ASC|   -1.59020***      .16806    -9.46  .0000    -1.91960  -1.26080

--------+--------------------------------------------------------------------

***, **, * ==>  Significance at 1%, 5%, 10% level.

Model was estimated on Mar 28, 2020 at 06:25:54 PM

-----------------------------------------------------------------------------



Error    805: Initial iterations cannot improve function.Status=3

Function value was   .1588384288D+04 at entry -----------

                     .3124865903D+05 at exit  -----------

Error   1025: Failed to fit model. See earlier diagnostic.


----
Mahi Puri

PhD Candidate and Graduate Assistant
Wildlife Ecology and Conservation
University of Florida