Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

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

Re: st: -svy: mean- with more than 400 subpopulations

From   [email protected] (Jeff Pitblado, StataCorp LP)
To   [email protected]
Subject   Re: st: -svy: mean- with more than 400 subpopulations
Date   Fri, 28 Jan 2011 18:18:31 -0600

Julian Reif <[email protected]> found a bug in Stata's -mean- command (also
present in -proportion-, -ratio-, and -total-) with two or more variables and
more than 400 subpopulations implied by the -over()- option:

> I am using the -svy: mean- command to calculate means for two different
> variables for 432 subgroups of a large survey and appear to be encountering
> a small bug.  It looks like Stata allows only 400 subgroups for the first
> variable, and then stores the 401-432 results under the name of the second
> variable twice (see output table below).  This is not just a display issue;
> typing -matrix list e(b)- at the prompt confirms that this is how the
> results are stored in the estimation matrices as well.  If I calculate the
> mean for only the first variable (i.e., -svy, over(group1 group2 group3):
> mean var1- instead of -svy, over(group1 group2 group3): mean var1 var2-)
> then the results are stored correctly.  

We have determined the cause and fixed it.

The fix should be available in the next ado-file update.

[email protected]
*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index