For "column" read "variable" throughout.
For "STATA" read "Stata" throughout.
The limits on matrix size are given in the help
on -limits-. The limits you quote apply to Intercooled
Stata, but with Stata/SE and Stata/MP you can go
much bigger.
But that is not relevant here. As Alex indicates,
you can use -summarize- directly.
So, for example,
foreach v of var a b c d {
su `v', meanonly
gen st`v' = (`v' - r(min)) / (r(max) - r(min))
}
Also, see the results of
findit std01
for the same idea implemented otherwise.
Nick
[email protected]
Alex Ogan
> There are several ways to do this. -search min- and -search
> max- would
> point you to them, although it wouldn't be immediately obvious which
> links to click on. Anyway, here are two:
>
> -summarize varname- will create scalars r(min) and r(max).
>
> -egen min_varname = min(varname)- will create a new variable
> containing
> the minimum across all observations.
Laura Flamand
> I am working in a program to estimate an index similar to the Index of
> Human Development (PNUD, various years), and I need to find the
> maximum and minimum value of several variables across 2500
> observations, I will then use these values to normalize the variables.
> I know how to find min and max values across rows, but I cannot find a
> command to find these min and max values across columns, other than to
> transpose the data set with xpose, or to transform my data set into a
> matrix (but then I can only have a 800x800 matrix, right?)
*
* 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/