estout with sem

[Erin]

I have successfully run a sem model, and also am able to output the table of results, at least for a single model.
I do not need to report the variance, covariances, and means of coefficients but cannot seem to drop them from the table.

so, this works:
estout, stats(aic bic N) cells(b(star fmt(3)) t(par fmt(2)))

but this doesn't:
estout, stats(aic bic N) cells(b(star fmt(3)) t(par fmt(2))) drop(_mean* _cov* _var*)

any ideas?

[Felix Leung]

Hi Erin,

It seems you want to suppress some of the equations. In the user-written estout command, you can do so with drop(equation_name: ). Without the colon, it would think it is the parameter name.

Here is an example:

use http://www.stata-press.com/data/r13/sem_1fmm

sem (x2 x3 x4 <- X)
eststo res
estout res, cells(b t) unstack

* to drop equations named "var(e.x2)" and "var(e.x4)"
estout res, cells(b t) unstack drop(var(e.x2): var(e.x4):)

* or use a wildcard to drop all "var(*)" equations
estout res, cells(b t) unstack drop(var(*):)

Here are the relevant outputs:

. estout res, cells(b t) unstack

-------------------------------------------------------------------------------------------------------
                      res                                                                              
                       x2           x3           x4    var(e.x2)    var(e.x3)    var(e.x4)       var(X)
                      b/t          b/t          b/t          b/t          b/t          b/t          b/t
-------------------------------------------------------------------------------------------------------
X                       1      .873456     5.858284                                                    
                        .     9.123494     10.28549                                                    
_cons            97.28455     97.09756     690.9837     94.99651     101.3808     342.8606     163.6298
                 67.09035     71.59739     99.27731     5.804173     6.576617     .9016638     5.021104
-------------------------------------------------------------------------------------------------------

. 
. * to drop equations named "var(e.x2)" and "var(e.x4)"
. estout res, cells(b t) unstack drop(var(e.x2): var(e.x4):)

-----------------------------------------------------------------------------
                      res                                                    
                       x2           x3           x4    var(e.x3)       var(X)
                      b/t          b/t          b/t          b/t          b/t
-----------------------------------------------------------------------------
X                       1      .873456     5.858284                          
                        .     9.123494     10.28549                          
_cons            97.28455     97.09756     690.9837     101.3808     163.6298
                 67.09035     71.59739     99.27731     6.576617     5.021104
-----------------------------------------------------------------------------

. 
. * or use a wildcard to drop all "var(*)" equations
. estout res, cells(b t) unstack drop(var(*):)

---------------------------------------------------
                      res                          
                       x2           x3           x4
                      b/t          b/t          b/t
---------------------------------------------------
X                       1      .873456     5.858284
                        .     9.123494     10.28549
_cons            97.28455     97.09756     690.9837
                 67.09035     71.59739     99.27731
---------------------------------------------------

[Roman Mostazir]

It seems the options for "estout" are not working (Stata-v.15.1):

use http://www.stata-press.com/data/r13/sem_1fmm, clear

sem (x2 x3 x4 <- X) // Output omitted

eststo res

estout res, cells(b t) unstack

equation / not found

[Ilana Akresh]

Do you know of any workaround for this? I am trying to output results from several multilevel multinomial logit models (as odds ratios) and it had been working with estout, but that is no longer the case. Roman Mostazir

[Roman Mostazir]

Not sure what is the source of the problem but in my case, it is the 'unstack' option that was causing the trouble. Again not sure why!

use http://www.stata-press.com/data/r13/sem_1fmm, clear

sem (x2 x3 x4 <- X) // Output omitted
------------------------------------------------------------------------------
             |                 OIM
             |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Measurement  |
  x2         |
           X |          1  (constrained)
       _cons |   97.28455   1.450053    67.09   0.000      94.4425    100.1266
  -----------+----------------------------------------------------------------
  x3         |
           X |    .873456    .095737     9.12   0.000     .6858149    1.061097
       _cons |   97.09756   1.356161    71.60   0.000     94.43953    99.75559
  -----------+----------------------------------------------------------------
  x4         |
           X |   5.858284   .5695676    10.29   0.000     4.741951    6.974616
       _cons |   690.9837   6.960138    99.28   0.000     677.3421    704.6254
-------------+----------------------------------------------------------------
    var(e.x2)|   94.99651   16.36693                      67.77262    133.1561
    var(e.x3)|   101.3808   15.41534                      75.25357    136.5791
    var(e.x4)|   342.8606   380.2533                      39.00168     3014.06
       var(X)|   163.6298   32.58842                      110.7485    241.7616
------------------------------------------------------------------------------


eststo m1

estout m1, cells(b t) unstack

equation / not found

estout m1, cells(b t)
-------------------------
                       m1
                      b/t
-------------------------
x2                       
X                       1
                        .
_cons            97.28455
                 67.09035
-------------------------
x3                       
X                 .873456
                 9.123494
_cons            97.09756
                 71.59739
-------------------------
x4                       
X                5.858284
                 10.28549
_cons            690.9837
                 99.27731
-------------------------
/                        
var(e.x2)        94.99651
                 5.804173
var(e.x3)        101.3808
                 6.576617
var(e.x4)        342.8606
                 .9016638
var(X)           163.6298
                 5.021104
-------------------------