st: Re: how to drop to get a balanced panel

["Michael Blasnik" <michael.blasnik@verizon.net>]

...
Here's one way:

egen nmiss=rmiss(bodyweight var2 var3 var4)
bysort id (nmiss): drop if nmiss[_N]>0


The first command counts the number of missing values across the variable list 
of interest for each observation.  The second command sorts by id and then 
number of missing values and drops the entire id if there are any missing values 
(by checking the maximum number of missing values for that id).

Michael Blasnik


----- Original Message ----- 
From: <west--@libero.it>
To: "statalist" <statalist@hsphsun2.harvard.edu>
Sent: Saturday, May 19, 2007 12:30 PM
Subject: st: how to drop to get a balanced panel



> Dear users,
>
> I have an unbalanced panel of students ("id" = 1, 2, ..., 5040), from 1998 to 
> 2002 ("time"), and a number of variables (ex. body weight...). I need generate 
> a complete panel dataset, by eliminating every student for which there is at 
> least a single item data missing for a year (or more). Could you help me with 
> an example? Thank you.
>
> Damiano Rossi
*
*   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/