想增加一个循环让它找到离event_day 最近的trading day

gen double d=abs(event_day-trading_day)
egen double x= min(d), by(group_id)
gen evt_dat_near_t_day =trading_day if x==d

?

Much thanks, will take a look.

I solve the problem like this:

gen absdif=abs(stockdate-event_date)
egen abstd=min(absdif),by(group_id)
replace dif=abstd-absdif if dif==. & event_date>=stockdate
replace dif=absdif-abstd if dif==. & event_date<stockdate

Hi, I met a new obstable when running the regression below, the error is shown as above.
The reason is, in some certain cases, the first row of the return is empty (something wrong with the data), but the security info of the first stockdate often would not be used in the regression, but stata shows "no observations".

How could i jump it or neglect it? I tried to drop id=194, but it does not work.

gen predicted_return=.egen id=group(company_id)  /* for multiple event dates, use: egen id = group(group_id) */forvalues i=1(1)N { /*note: replace N with the highest value of id */         l id company_id if id==`i' & dif==0        reg ret market_return if id==`i' & estimation_window==1         predict p if id==`i'        replace predicted_return = p if id==`i' & event_window==1         drop p}