[Limdep Nlogit List] ";Export" issue

Suzanne Clain suzanne.clain at villanova.edu
Sat Mar 12 10:23:07 AEDT 2016

Matt --

Did you notice that it appears to be exporting the interval information presented in NLOGIT (i.e. all six columns of numerical values per variable) and misaligning that information in your csv output (which only has four columns of numerical values per variable)?

I don't know if you can do anything to control that or if there's an error in the programming of the export command.

From: limdep-bounces at limdep.itls.usyd.edu.au [limdep-bounces at limdep.itls.usyd.edu.au] on behalf of Matt Quaife [matthew.j.quaife at gmail.com]
Sent: Thursday, March 10, 2016 6:44 PM
To: Limdep at limdep.itls.usyd.edu.au
Subject: [Limdep Nlogit List] ";Export" issue

Hi all,

I am having some trouble applying the ;export command, specifically the
resultant .csv file contains coefficients and standard errors etc. that are
(sometimes very) different to those presented in NLOGIT.

I am using the syntax below:


;lhs=choice, cset,alti
;tree=switch(a,b,c), noswitch(n)
;export output
etc $

In the exported file, the first attribute information appears to be
correct, however the second coefficient onwards exports incorrect
information. Example below (please excuse formatting, I am not a listserv

NLOGIT output:
        |                         Standard            Prob.      95%
  CHOICE|  Coefficient       Error       z    |z|>Z*         Interval
        |Attributes in the Utility Functions (beta)
     Att1|    -.01437         .24718     -.06  .9537     -.49884    .47011
     Att2|     .07179         .17488      .41  .6814     -.27097    .41455
     Att3|     .12039         .29830      .40  .6865     -.46427    .70505

Excel .csv output:
Variable Coeff.   Std.Err. t-ratio  P-value
Att1     -1.44E-02 0.247185 -5.81E-02 0.953656
Att2      -0.498839 0.470108 7.18E-02 0.17488
Att3     0.410522 0.681423 -0.270966 0.41455

Advice much appreciated.

Best wishes
Limdep site list
Limdep at limdep.itls.usyd.edu.au

More information about the Limdep mailing list