Ada:
Regression commands store the variance covariance matrix. The standard errors of the coefficients are the square roots of the diagonal elements of that matrix, and that is what the line -mat se= vecdiag(cholesky(diag(vecdiag(e(V)))))'- does. Now, you can also just store the diagonal elements and take the square root ones you have put them in a dataset, like in the example below. This should not break down when zeros occur.
HTH,
Maarten
*-------------begin example-----------
clear
sysuse auto
regress mpg len turn head gear
qui{
drop _all
mat b=e(b)'
mat var= vecdiag(e(V))'
gen vars=""
local names: rownames b
tokenize `names'
local num : word count `names'
set obs `num'
forvalues i=1/`num'{
replace vars= "``i''" in `i'
}
svmat b
svmat var
rename b1 coef
gen se = sqrt(var1)
drop var1
gen t=coef/se
gen pvalue=2*ttail(e(df_r),abs(t))
gen lb=coef-invttail(e(df_r),0.025)*se
gen ub=coef+invttail(e(df_r),0.025)*se
}
list
*-----------------end example--------------------
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting adress:
Buitenveldertselaan 3 (Metropolitan), room Z214
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
-----Original Message-----
From: [email protected] [mailto:[email protected]]On Behalf Of Ada Ma
Sent: dinsdag 18 april 2006 17:09
To: [email protected]
Subject: Re: st: RE: how do i save the significance levels of coefficients after estimation?
Thanks for the suggestion. I'm looping the estimation through
different samples, so often quite a (big) bunch of the variables got
dropped. I'll need to think of a way to drop the variables or the
zeros from the matrices.
Cheers,
Ada
On 4/18/06, Maarten Buis <[email protected]> wrote:
> Probably the easiest way to do that is to not enter the dropped the variables in the regression statement, and use the e(V) of that regression.
> HTH,
> Maarten
>
> -----------------------------------------
> Maarten L. Buis
> Department of Social Research Methodology
> Vrije Universiteit Amsterdam
> Boelelaan 1081
> 1081 HV Amsterdam
> The Netherlands
>
> visiting adress:
> Buitenveldertselaan 3 (Metropolitan), room Z214
>
> +31 20 5986715
>
> http://home.fsw.vu.nl/m.buis/
> -----------------------------------------
>
> -----Original Message-----
> From: [email protected] [mailto:[email protected]]On Behalf Of Ada Ma
> Sent: dinsdag 18 april 2006 16:40
> To: [email protected]
> Subject: Re: st: RE: how do i save the significance levels of coefficients after estimation?
>
> the estimation I did have lost of
> dropped variables, and when I tried to implement:
>
> mat se= vecdiag(cholesky(diag(vecdiag(e(V)))))'
>
> Stata replies saying:
> matrix not positive definite
> r(506);
>
> How can I trim the zeros from the e(b) and e(V) matrices before I
> perform the commands posted by Maarten?
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
