RE: st: Maximum value for a variable

[Nick Cox <n.j.cox@durham.ac.uk>]

Note that even in principle the code here will probably not work. 

Unless a local macro `ayear' has been defined, the second time around the loop the 
-egen- call will fail as the variable to be created already exists. 

I agree with Oliver that -rolling- (or as said -mvsumm- from SSC) is a better idea. 

Nick 
n.j.cox@durham.ac.uk 

Oliver Jones

maybe you could also try something manual like:

sort country year
forvalues to_year = 1999/2010{
	local from_year = `to_year' - 14
	by country year: egen max_growth`ayear' = max(gdp_growth) ///
		if year > `from_year' & year < `to_year'
}

But I guess rolling is more convenient...

Am 22.03.2011 18:13, schrieb ajjee:

> I have a panel dataset over the period 1960-2010 for 30 countries. I want to
> compute the maximum value for a variable over the last 15 years for each
> country. For example, for Austria, I want to compute the max of gdp growth
> from 1984-1999 for the year 1999 and 1985-2000 for the year 2000 and so on.

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