[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
"Ergo, Alex" <aergo@jhsph.edu> |

To |
"statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |

Subject |
st: Extracting the largest possible common set of variables for a given number of surveys |

Date |
Wed, 26 Nov 2008 14:34:40 -0500 |

Dear all, I have data from a large number of surveys (150). There are about 300 different variables, but not all of these variables are available in all the surveys. I've entered the information on data availability in a STATA dataset as follows: var001 var002 var003 var004 var005 ... var299 var300 dataset001 1 1 . . 1 . . dataset002 1 . 1 . . 1 . dataset003 . 1 . . 1 . 1 ... dataset150 1 1 1 1 1 1 1 (where the value is "1" if the variable is available, "." otherwise) I'd like to be able to determine, for a given number of surveys, which combination of surveys I should select in order to have the largest number of variables in common. For examples, if I decide to include 15 surveys in my analysis, which ones should I select to have the maximum number of variables (available in all 15 surveys) and what are these variables? What if I decide to include 16 surveys in my analysis? etc... Unfortunately, the command mvpatterns doesn't work with such large number of variables. The command misschk doesn't do the job either. Can anyone think of another way to extract this information using STATA? Thanks, Alex * * 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/

- Prev by Date:
**st: Question about options of "xtreg, re"** - Next by Date:
**Re: st: Question about options of "xtreg, re"** - Previous by thread:
**st: Question about options of "xtreg, re"** - Next by thread:
**st: Using xtgee with multiple weights per ID** - Index(es):

© Copyright 1996–2015 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |