significance stars with coefficients in matrix

[Christian Vienna]

Hi all,

I run a set of regressions, where the independent variable as well as the lag of the independent variable vary, so that I obtain 10x4 regressions. For matters of overview I put the obtained coefficients into a matrix. I also put the standard errors into a second matrix.
However, I would much prefer to have significance asterisks next to the coefficients, for obvious oversight reasons.

Unfortunately I have not yet figured out how to do this.

my command so far is:
I only include the part of the code that I think is crucial for my question. below is the extended code:

matrix beta_01 = J(4,10,.)
matrix se_01 = J(4,10,.)

 matrix beta_01[`i'+1,k] = _b[myindep`var'_lag`i']
 matrix se_01[`i'+1,k] = _se[myindep`var'_lag`i']

In this version of my code, I included my actual regression code, even though I don't think the regression matters. But mybe it helps:

local myvarlist 10 vars
matrix beta_01 = J(4,10,.)
matrix se_01 = J(4,10,.)
foreach var of varlist `myvarlist' {
scalar m =m+1
    qui regress mydepvar myindep`var'
    matrix beta_01[1,m] = _b[myindep`var']
    matrix se_01[1,m] = _se[myindep`var']
scalar k = k+1
forval i=1/3 {
qui regress mydepvar myindep`var'_lag`i'
 matrix beta_01[`i'+1,k] = _b[myindep`var'_lag`i']
 matrix se_01[`i'+1,k] = _se[myindep`var'_lag`i']    
}
}

Kind regards
Christian

[Phil Bromiley]

The easy way is to use outreg2 or estout which are all set up do to the stars. Programming this seems like a lot of trouble when the software is already available.

As the User Guide says, "to Stata, a matrix is a named entity containing an r c (0 < r matsize, 0 < c matsize) rectangular array of double-precision numbers (including missing values) that is bordered by a row and a column of names". So, you can't put stars in a matrix. If you are determined to program this yourself, you could write into variables.