mixlogit: choose different base alternative

[Joost Verlaan]

I am modelling choices using the -asmixlogit- command.
I want to have a different base alternative.
The output of the following command,

asmixlogit choice td, case(group) alternatives(alternative) random(at fee, correlated) corrmetric(cholesky) casevars(oint) iter(5)

is as follows:

Alternative-specific mixed logit               Number of obs      =      4,728
Case variable: group                           Number of cases    =      1,576

Alternative variable: alternative              Alts per case: min =          3
                                                              avg =        3.0
                                                              max =          3
Integration sequence:      Hammersley
Integration points:               100             Wald chi2(5)    =      71.73
Log simulated likelihood = -1089.7482             Prob > chi2     =     0.0000

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
alternative  |
          td |  -.1235769   .0359336    -3.44   0.001    -.1940055   -.0531482
          at |  -.1918075   .0574084    -3.34   0.001    -.3043259    -.079289
         fee |  -8.469227   1.086547    -7.79   0.000    -10.59882   -6.339633
-------------+----------------------------------------------------------------
Normal       |
        l(at)|   .3164573   .1010057                      .1692902    .5915596
    l(at,fee)|  -1.979668   .5795327    -3.42   0.001    -3.115531   -.8438048
       l(fee)|   .5465202    .876635                      .0235642    12.67533
-------------+----------------------------------------------------------------
FSP          |  (base alternative)
-------------+----------------------------------------------------------------
PSP          |
        oint |  -.1036039   .3159861    -0.33   0.743    -.7229252    .5157174
       _cons |   2.649625   .6643864     3.99   0.000     1.347452    3.951799
-------------+----------------------------------------------------------------
PUP          |
        oint |  -.0986893   .3154458    -0.31   0.754    -.7169518    .5195731
       _cons |   6.835244   1.109541     6.16   0.000     4.660584    9.009904

So, basically I want to have PSP as base alternative instead of FSP.
My alternative variable is FSP, PSP, PUP, so I think that Stata just picks the first one as the base alternative. I can try to change this of course and see how that works out, but I would like to know how to this is in Stata as well.

[Andrew Musau]

For any command that you deal with in Stata, you can always type help [command name] and chances are that you will find the relevant information.

help asmixlogit

yields

basealternative(#|lbl|str) alternative used for normalizing location

which implies that you just need to add

 smixlogit choice td, case(group) alternatives(alternative) basealternative(PSP)...

or the value of the alternative.