I need to bootstrap a large number of statistics, actually, it is 5
statistics, calculated for the different year-category groups of the
data set. Up to now I have no luck in obtaining bootstrap estimates for
the number of statistics larger than 4.
I use the following program, which works fine for 4 statistics :
program define jobflo
version 6.0
if "`1'" == "?" {
global S_1 "POS_98 NEG_98 POS_99 NEG_99"
exit
}
local i=1998
while `i'<=1999 {
summarize pos if year == `i'
local bf`i' = r(sum)
summarize neg if year == `i'
local cd`i' = r(sum)
summarize xit if year == `i'
local ef`i' = r(sum)
local gh`i' = `bf`i' ' / `ef`i''
local jk`i' = `cd`i' '/`ef`i''
local i = `i'+1
}
post `1' `gh1998' `jk1998' `gh1999' `jk1999'
'