Using spkde for subsets of the data

[John MacDonald]

I would like to generate a kernel density estimate at the year-month (ym) level.
I tried forval but I get the exact same values for each year-month. Can anyone see if there is something wrong with my code?

forval i=750/773 {
spkde using "PHL-GridPoints.dta" if ym==`i', ///
xcoord(_X) ycoord(_Y) ///
bandwidth (fbw) fbw(1000) dots ///
edgecorrection ///
noverbose saving ("Kde_gun`l'.dta", replace)
}