Re: st: Re: Exclude non-positive values when using collapse

["Joseph Coveney" <stajc2@gmail.com>]

Chris Yang wrote:

In my example, I collapse by group, so that for group 1 there are only
four values for var3: 1, 1, -1, 1. If we exclude the -1, the mean
after collapse is (1 + 1 + 1) / 3.

If I understand you correctly, after replacing -1 values with missing
values, the -collapse- command will simply ignore the missing ones
when averaging for variables. Is that correct?

--------------------------------------------------------------------------------

Sorry, I skipped over the by-group part by accident. As to your question, here's
one way to find out.

Joseph Coveney

. input byte(var1 var2 var3 group)

         var1      var2      var3     group
  1.  0     1      1     1
  2.  0     -1     1     1
  3.  1     0      -1    1
  4.  1     1      1     1
  5.  1     1      1     2
  6.  1     0      0     2
  7. end

. mvdecode var?, mv(-1)
        var2: 1 missing value generated
        var3: 1 missing value generated

. collapse (mean) var?, by(group)

. list, noobs

  +-------------------------------+
  | group   var1      var2   var3 |
  |-------------------------------|
  |     1     .5   .666667      1 |
  |     2      1        .5     .5 |
  +-------------------------------+

. exit

end of do-file

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/