data have; input id day date :yymmdd10. drink age; format date yymmdd10.; cards; 1 1 20180907 1 24 1 2 . 0 24 1 3 20180909 3 24 2 1 20171211 2 30 2 2 . 0 30 2 3 . 0 30 ; data temp; set have(where=(date is not missing)); by id; if first.id; keep id day date; run; data want; merge have temp(rename=(day=_day date=_date)); by id ; if missing(date) then date=_date-(_day-day); drop _:; run; ... View more

If you do this the easy way, you don't even need to know the variable names.  You just need data that is in sorted order by ID:   data want; update have (obs=0) have; by id; output; run; ... View more

Thank you for the reference reading! I got that "want" is the newly modified data set by revising master data set "have(obs=0)" by the information in the data set "have". Could you help explain what does "have (obs=0)" refer to? Obs is not a variable right, so what does obs=0 refer to? ... View more