xtpmg command stata

[Sean O'Connor]

Hi folks!

I'm looking to use the xtpmg command on STATA, however, I'm running into a problem. When I try to run the command I get the following error

matrix operation not found
r(501);

Have I excluded something from the command by chance?

Sean

[Nick Cox]

 
help set trace

to learn about showing us where an error occurs. Also, please study FAQ Advice, including Sections 12 (showing the exact code used; explaining user-written commands) and 18 (how to avoid an error that would embarrass you in presentations, dissertations or job applications).

[Richard Williams]

Since you don't show us what you typed it is hard to say. But it works fine with the xtpmg_examples.do file that you can download with the program (available from SSC). I suggest you check those examples and if it isn't obvious that you are doing something wrong. then post your code and output.

A program by that name is available from both SJ and SSC. I don't know if they are different or not. You should indicate where user-written commands come from. This is especially important when more than one version of the program exists. The version you are using might have bugs that were fixed in the other version. When multiple versions of a program exist I think the SSC version tends to be the most current but there are some noteworthy exceptions to that, e.g. SSC has assorted programs by Long & Freese that are NOT the most current versions (which Long now keeps on his own personal web site.)

[Sean O'Connor]

I'm still having trouble with these commands, particularly in regards to interpreting them.

 xtpmg  D_rgdpl D_ki D_oilp if year>=1985, lr( log_rgdpl log_ki log_oilp) ec(ec) replace pmg
  Iteration 0:   log likelihood =  3589.2148  (not concave)
Iteration 1:   log likelihood =    3591.55  (not concave)
Iteration 2:   log likelihood =  3591.6188  (not concave)
Iteration 3:   log likelihood =  3604.1975 
Iteration 4:   log likelihood =  3613.2808 
Iteration 5:   log likelihood =   3613.903 
Iteration 6:   log likelihood =  3613.9722 
Iteration 7:   log likelihood =  3613.9776 
Iteration 8:   log likelihood =  3613.9776 
  Pooled Mean Group Regression
(Estimate results saved as pmg)
  Panel Variable (i): id                          Number of obs      =      1584
Time Variable (t): year                         Number of groups   =        66
                                                Obs per group: min =        24
                                                               avg =      24.0
                                                               max =        24
                                                  Log Likelihood     =  3613.978
------------------------------------------------------------------------------
     D_rgdpl |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
ec           |
      log_ki |   4.606324   .8481577     5.43   0.000     2.943965    6.268683
    log_oilp |   2.467857   .5870011     4.20   0.000     1.317356    3.618359
-------------+----------------------------------------------------------------
SR           |
          ec |   .0012516   .0033083     0.38   0.705    -.0052326    .0077357
        D_ki |   .1341978   .0223032     6.02   0.000     .0904844    .1779113
      D_oilp |   .0852191   .0574677     1.48   0.138    -.0274155    .1978537
       _cons |   .0406241   .0418264     0.97   0.331    -.0413542    .1226024
------------------------------------------------------------------------------

Would anyone be kind enough to walk through how to interpreted these kinds of output? I'm generally more use to OLS estimations etc. Also when I try to run other analysis, such as

 xtpmg D_rgdpl D_ki D_oilp D_reserves D_exports D_price PR free partiallyfree notfree if year>=1986, lr( log_rgdpl log_ki log_oilp
> log_reserves log_exports log_price PR free partiallyfree notfree) ec(ec) replace mg op. sys. refuses to provide memory
r(909);

I've increased the mat size to the maximum and decrese the amount of variables to two in some cases but the same error keeps appearing. Anyone with any idea how this could be resolved?