Roy,
thank you for your reply and help.
-anymatch- does a good work. However, maybe there is still a problem.
In my posting, I need to find a matching observation from the large sample for each of the object sample. Following -anymatch- may result in one observation in the object sample matches the other one also from the obejct sample, which is not what I need.
Hope for any help.
Best regards,
Rose.
----- Original Message -----
From: Roy Wada <[email protected]>
To: [email protected]
Subject: Re: st: How to find a matching sample for a specified sample based on some standards?
Date: 2009-10-13 18:51:21
> I need to find a matching sample for my object sample.
> Its matching sample is chosen from a large sample based on some filtering standards that
> the object and its matching should be 爄n the same year, the same industry and the asset of
> the matching should 爁luctuate at the 30% upper or lower than the object. If there are more than one
Try -anymatch- from ssc.
The wrinkle is your having to do it within the same year and industry.
In that case, keep within each group and append the result.
egen group=group(year industry)
save temp, replace
qui sum group
local max=`r(max)'
forval num=1/`max' {
use temp, clear
keep if group==`num'
anymatch asset, id(id) metric(asset)
save mystuff`num', replace
}
... and append those files together here ....
The distance will be automatically produced. Keep/drop the ones within
the 30% limit.
Roy
*
* 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/
*
* 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/