st: Mata : getting covariance matrix with moptimize()

[To Maxime <Maxime.To@ensae.fr>]

Hi,

I am using mata to estimate moment based estimators. To do so I use the moptimize() function with a Gauss-Newton method (gn). 

When I compare to the use of the gmm()  function, results are correct but I can't obtain vce results. I give you here a simplifyed version of my code that has the same problem.

mata:
 mata clear
 function i_gmm(M,todo,b,r,S)
  {
  y1 = moptimize_util_depvar(M, 1)
  x1 = moptimize_util_indepvars(M, 1)
  p1 = moptimize_calc_xb(M, b, 1)
  r  = x1'p1
  }
  

 W= st_matrix(W)

 M = moptimize_init()
 moptimize_init_evaluator(M, &i_gmm())
 moptimize_init_which(M,"min")
 moptimize_init_technique(M, "gn")
 moptimize_init_evaluatortype(M, "q0")
 moptimize_init_depvar(M, 1, "devar")
 moptimize_init_eq_indepvars(M, 1, "indepar1 indepvar2")
 moptimize_init_gnweightmatrix(M,W)
 moptimize_query(M) 
 moptimize(M)
 moptimize_result_display(M)
end

Thanks,
Maxime
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/