Re: st: Why can't I egen means with pweights?

[n j cox <n.j.cox@durham.ac.uk>]

Laura Schechter wrote:

>    I have been trying to program a specific bootstrap in stata, and need
> to use the weighted means of certain variables over time. Unfortunately,
> it seems to me that the egen command doesn't work with weights. For now, I
> have just been using the series of commands
>    svymean varname [pweight=x], by(month)
>    matrix e=e(est)
>    generate ET=el(e,1,1) if month==1
>    replace ET=el(e,1,2) if month==2
>       etc. etc. etc.
> But this is really slow, and it would be so convenient if the egen command
> would work with weights so I could just say something like
>     egen ET=mean(varname) [pweight=x], by(month)
> Does anyone know a way of doing this??
The answer to your title question is: because weights
were never part of the syntax of -egen-. I don't know why.

An answer to your end question is to program it
directly. A similar problem is documented at
"I want to calculate a variable containing weighted group summary 
statistics, but I do not want to collapse the data and egen does not 
support weights. How can I do this?"
http://www.stata.com/support/faqs/data/weighted.html

In your case -levels- is presumably unnecessary if -month-
takes on values 1(1)12.

Hence

gen ET = .
qui forval i = 1/12 {
	su varname [pweight=x] if month == `i', meanonly
	replace ET = r(mean) if month == `i'
} 	

Nick
n.j.cox@durham.ac.uk

*
*   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/