Combining svy SEs with estat sd output and extracting it out of stata?

[Mehmet Ozmen]

Hello StataList-ers!

I've used Stata on and off and have given it a real go recently, and its been fantastic but I've had a difficulty getting results out.
I have my survey all set up and want to create weighted means, standard errors, and standard deviations (yes, weighted sd, you heard that right, its for population norms)

So...

svy: mean YVAR, over(agegroup)
estat sd

and presto I have two tables, first with mean and stand. errors and the second mean and standard deviation. All the information is correct.
Now, I'm trying to find a way of calling the mean and standard error from the first and appending the standard deviation from the second and creating a third table as my output.

I've played with esttab, estout and outreg2 and I have't quite managed it. I have to do this 144 times with different variables and over() categories.
I thought I came the closest with esttab but no luck.

Any suggestions?

[Aspen Chen]

It might be possible, but I don't know how to conveniently combine e-class estimates with a r-class matrix. My preferred solution to this type of problems is usually a plain matrix. For example:

svy: mean YVAR, over(agegroup)
estat sd
mat b=e(b)
mat V=e(V)
mat sd=r(sd)
loc n=rowsof(V)
mat est=J(`n',3,.)
mat colnames est=b se sd

forval i=1/`n'{
  mat est[`i',1]=b[1,`i']
  mat est[`i',2]=sqrt(V[`i',`i'])
  mat est[`i',3]=sd[1,`i']
}
mat list est

Does this address your issues? A small function would be useful to repeat this process.