Steven Fraser's problem can also be approached using standard
Stata commands (although I was interested to see a demonstration of -ineq-)
> Date: Thu, 13 Jun 2002 09:48:59 -0400
> From: "Fraser, Steven" <[email protected]>
> Subject: st: Tab and Matcell
>
> Dear Statalist:
>
> I am having a problem implementing matrix operations within the 'tab'
> command. Using a variation of the familiar car example, assume I have the
> following dataset:
>
> Manuf Model Class MPG
> ford explorer SUV 12
> ford expedition SUV 10
> ford focus compact 34
> dodge durango SUV 16
> gm tahoe SUV 15
>
> Assume I have hundreds of automobiles (cars, trucks, and SUVs).
> I am trying
> to develop a focus measure (a Herfandahl index) that captures the focus of
> the manufacturer.
bysort manuf class : gen i_2 = _N if _n == 1
replace i_2 = i^2
bysort manuf : gen n_2 = _N if _n == 1
replace n_2 = n^2
egen simpson = sum(i_2) , by(manuf)
replace simpson = simpson/n_2
li manuf simpson if simpson ~= .
by manuf : replace simpson = simpson[1]
This longer approach has the (small) advantage over -ineq-
that the index is added to the existing data base, not to the collapsed
version, so can be compared at once with any other variables of interest.
************************
Paul Seed
[email protected]
Medical Statistician
Wolfson Institute of Preventive Medicine
Charterhouse Square
London EC1M 6BQ
(020) (+44) 7882 6298
fax 7882 6270
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
