Re: st: Conditional Variable means to new observation

[Maarten Buis <maartenlbuis@gmail.com>]

On Fri, Aug 30, 2013 at 3:05 PM, Nickolas Lyell wrote:
> I am looking to take a conditional mean (or sum) of a variable and include it as a new observation.
>
> For instance, I have data with several county indicators horizontally and county ids vertically.  I would like to take the mean growth rate (a variable) for only those counties that are Large (LgMdSm==2) and create a new observation that contains that value under the variable growth rate.

You almost never want to store those numbers as an extra row in your
data. Stata takes the definition of a dataset very strictly, and
rightly so: the rows are the units and the columns are characteristics
of those units. All large counties together does not represent a new
unit. However, that mean growth rate you want to compute is a
characteristic shared by all counties that are "large": so that mean
has to be stored as a column. Here are two ways of computing such
means:

*------------------ begin example ------------------
// create some example data
clear
set obs 10
gen county_id = _n
gen LgMdSm = (_n > 5) + 1
gen growth = rnormal()

// first method
egen mean_growth = mean(growth) if LgMdSm == 2

// second method
bys LgMdSm : egen mean_growth2 = mean(growth)

// see the results
list
*------------------- end example -------------------
* (For more on examples I sent to the Statalist see:
* http://www.maartenbuis.nl/example_faq )

Hope this helps,
Maarten

---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany

http://www.maartenbuis.nl
---------------------------------

*
*   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/