How do I calculate least square means in Stata?

Currently there is no convenience command in Stata to calculate the least square means, but one may use the adjust command to compute them manually. LSMEANS are just the predicted linear combination xb while holding the other covariates at values 1/n, where n is the number of categories in the corresponding discrete covariates. For continuous covariates, one just holds their values at their means when making the prediction.

Following is an example using the auto dataset:

 . sysuse auto, clear
 (1978 Automobile Data)
 
 . anova mpg foreign rep78
 
                            Number of obs =      69     R-squared     =  0.2825
                            Root MSE      = 5.16246     Adj R-squared =  0.2256
 
                   Source |  Partial SS    df       MS           F     Prob > F
               -----------+----------------------------------------------------
                    Model |  661.189524     5  132.237905       4.96     0.0007
                          |
                  foreign |  111.773747     1  111.773747       4.19     0.0447
                    rep78 |  179.189006     4  44.7972516       1.68     0.1655
                          |
                 Residual |  1679.01337    63  26.6510059   
               -----------+----------------------------------------------------
                    Total |   2340.2029    68  34.4147485   
 
 . xi: regress mpg i.foreign i.rep78
 i.foreign         _Iforeign_0-1       (naturally coded; _Iforeign_0 omitted)
 i.rep78           _Irep78_1-5         (naturally coded; _Irep78_1 omitted)
 
       Source |       SS       df       MS              Number of obs =      69
 -------------+------------------------------           F(  5,    63) =    4.96
        Model |  661.189524     5  132.237905           Prob > F      =  0.0007
     Residual |  1679.01337    63  26.6510059           R-squared     =  0.2825
 -------------+------------------------------           Adj R-squared =  0.2256
        Total |   2340.2029    68  34.4147485           Root MSE      =  5.1625
 
 ------------------------------------------------------------------------------
          mpg |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
 -------------+----------------------------------------------------------------
  _Iforeign_1 |   3.556584   1.736681     2.05   0.045     .0861047    7.027064
    _Irep78_2 |     -1.875   4.081284    -0.46   0.648     -10.0308    6.280795
    _Irep78_3 |  -1.922325   3.774126    -0.51   0.612    -9.464315    5.619665
    _Irep78_4 |  -1.111626   3.944633    -0.28   0.779    -8.994345    6.771094
    _Irep78_5 |   3.453704   4.215132     0.82   0.416    -4.969565    11.87697
        _cons |         21   3.650411     5.75   0.000     13.70524    28.29476
 ------------------------------------------------------------------------------
 
 . adjust _Irep78_2=.2 _Irep78_3=.2 _Irep78_4=.2 _Irep78_5=.2, by(foreign) se
 
 -------------------------------------------------------------------------------
      Dependent variable: mpg     Command: regress
     Variable left as is: _Iforeign_1
 Covariates set to value: _Irep78_2 = .2, _Irep78_3 = .2, _Irep78_4 = .2, 
                          _Irep78_5 = .2
 -------------------------------------------------------------------------------
 
 ----------------------------------
  Car type |         xb        stdp
 ----------+----------------------
  Domestic |     20.709   (1.04909)
   Foreign |    24.2655   (1.55104)
 ----------------------------------
      Key:  xb    =  Linear Prediction
            stdp  =  Standard Error

There are five outcomes in the variable rep78. The values of the dummy variables _Irep* are held as 1/5 = 0.2.

Given that adjust is just another form of predict, we could also use predict to reproduce the results.