Dear all,

Assume that one runs a model using the bayes prefix. This model includes also a large number of indicators apart from some standard controls.

I am wondering if there is a way not to have the results of the indicators shown at all in the Stata output.

I will will present a minimal working example.

Code:
input id    year    y    x1    x2
1    1990    0.765290722    0.786241262    0.783572146
1    1991    0.611380163    0.739519495    0.60693537
1    1992    0.590601317    0.490306934    0.479196734
1    1993    0.708316766    0.97180118    0.40444494
1    1994    0.587700824    0.442869206    0.399139847
1    1995    0.092595406    0.475633824    0.055023827
2    1990    0.995706082    0.356060781    0.242968298
2    1991    0.688079045    0.36475865    0.829626742
2    1992    0.564536023    0.219309875    0.59165049
2    1993    0.728154736    0.370715204    0.319482427
2    1994    0.972550408    0.331836688    0.495801468
2    1995    0.907903102    0.643605887    0.04147437
3    1990    0.666757644    0.604741524    0.892619352
3    1991    0.945322031    0.076470116    0.83826142
3    1992    0.262225766    0.622579231    0.738801605
3    1993    0.14160545    0.420033382    0.883489729
3    1994    0.674649086    0.287176323    0.798404419
3    1995    0.462505639    0.782454944    0.206593929
4    1990    0.103318329    0.860360564    0.384564281
4    1991    0.892686701    0.387916868    0.911908676
4    1992    0.23782402    0.403880512    0.962117085
4    1993    0.733468764    0.376152156    0.237543589
4    1994    0.925702313    0.438672271    0.103990407
4    1995    0.309275933    0.771466795    0.119082256
end
Suppose we run the following model:

Code:
bayes, rseed(12345) : reg y x1 x2 i.id i.year
The results we get are:

Code:
Bayesian linear regression                       MCMC iterations  =     12,500
Random-walk Metropolis-Hastings sampling         Burn-in          =      2,500
                                                 MCMC sample size =     10,000
                                                 Number of obs    =         24
                                                 Acceptance rate  =       .327
                                                 Efficiency:  min =    .003156
                                                              avg =    .006709
Log marginal likelihood = -74.144851                          max =     .02003
 
------------------------------------------------------------------------------
             |                                                Equal-tailed
             |      Mean   Std. Dev.     MCSE     Median  [95% Cred. Interval]
-------------+----------------------------------------------------------------
y            |
          x1 |  .0390891   .1586424   .025329   .0357585  -.2741725   .3541185
          x2 | -.3099317   .2938798   .039644  -.3094318  -.8912901   .2766925
             |
          id |
          2  |  .2899881   .1213703   .021604   .2966582   .0641597   .5389639
          3  |   .055742   .1603441   .022674   .0559719  -.2698375   .3579688
          4  | -.0549487   .1484017   .016786  -.0567729  -.3385171   .2308046
             |
        year |
       1991  |  .2982491   .1879188   .026803   .3004807  -.1168325   .6647947
       1992  | -.1464675   .1949344   .020653  -.1421357  -.5566844    .232561
       1993  | -.0215075   .1920174   .031623  -.0132725  -.4371793   .3662122
       1994  |  .1843418   .1978521   .020934    .188863  -.2352899   .5520474
       1995  | -.2889066   .2437549    .03487  -.2840871  -.7785481   .1575815
             |
       _cons |  .6705034   .2512025   .040967   .6678414   .1940561   1.183284
-------------+----------------------------------------------------------------
      sigma2 |  .0759295   .0308258   .002178   .0698249   .0363304   .1570932
------------------------------------------------------------------------------
Is it possible to show (or save) only the results for x1 and x2? It appears that estimates table does not work with Bayes.

Thank you.