st: Splitting a file up into years, running analysis, and combining file again. Necessary due to a glitch in Stata command.

[Matthew Krauchunas <krauchunasms@mymail.vcu.edu>]

Hello,

I have data that I am working with for years 2002-2004 & 2006-2008.  I
am geocoding and creating alternative markets using a 15 mile and 25
mile radius via the user created CIRCNUM command.  However, I came
across a problem with the CIRCNUM command when I tried to run it by
groups (i.e., year).

* Creating nursing home HHI using a 15 mile radius *
by year: circnum localTotalSize15, radius(15) miles sum(lic_bed_month_avg)
by year: circnum nh_hhi15, radius(15) miles
sum("(lic_bed_month_avg[j]/localTotalSize15[i])^2")


It appears that it does not reset when it move on to the next group.
For instance, I segregated the data into three files:  2002, 2003, and
2002 & 2003 files.  The numbers for the 2002/2003 combined file do not
match the numbers when the years are run by themselves.  I have
notified the command's developer, but I cannot wait for a fix action.
Thus, my only recourse is to break the file up into individual years,
run the appropriate analyses using CIRCNUM, and then append the files.
 Does anyone have any suggestions?

My file name is G:\Merged_Joined_Appended files\clean oscar oshpd
merge wgis.dta.

Thanks for your help.

Matt
*
*   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/