Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Mitch Downey" <pmdowney@ucsd.edu> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: Selecting variables based on variable labels |
Date | Fri, 7 Sep 2012 18:09:20 -0700 (PDT) |
Hello, I'm using the National Longitudinal Survey of Youth. There are several series of variables where the same question was asked over multiple years. E.g., R0188000 TYPE OF RESIDENCE R IS LIVING 79 R0402800 TYPE OF RESIDENCE R IS LIVING 80 R0612100 TYPE OF RESIDENCE R IS LIVING 81 R0828400 TYPE OF RESIDENCE R IS LIVING 82 R1075700 TYPE OF RESIDENCE R IS LIVING 83 R1451400 TYPE OF RESIDENCE R IS LIVING 84 R1798600 TYPE OF RESIDENCE R IS LIVING 85 R2160200 TYPE OF RESIDENCE R IS LIVING 86 R2369100 TYPE OF RESIDENCE R IS LIVING 87 where R0188000 is the variable name and TYPE OF RESIDENCE R IS LIVING 79 is the variable label. Unfortunately, the variable names are random. This makes it impossible to select a group of variables using res* (for instance) or res79-res87. The only way I can find this group of variables is by typing -- lookfor "TYPE OF RESIDENCE R IS LIVING" -- I would like to identify a list of variables using the lookfor command and either create a list or rename them in a sequential way. Is there a way to do this? Can you build a variable list from lookfor results? My ultimate goal is to make the dataset long format, so if anyone has and idea to shortcut straight to that, I'm all ears. Thank you, Mitch * * 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/