@Greek wrote: Thank you very much for your quick response!    Many apologies, I tried posting a picture of the error message, but my company has a very conservative firewall that blocked it. There was no output, SAS EG just disconnected.     Logs are text, so copy text and paste text.   If the error is from the OS and not the log, which seems possible in this case, it may be that you are exceeding the output limits of your job. Having such large log entries may be an indication of other issues.   I suggest also using a subset of the SAMPSIZE data set so that you are only selecting for a couple of strata and examine the log to see why it is generating so much output. ... View more

You can not execute the LAG function conditionally. The LAG function has to be executed for every observation.   See the sample below with the modfied logic. You can make use of BY group processing and the FIRST. functionality.   data have; infile cards dlm=","; input count_cluster Max_salary ; cards; 1, 56768 1,70741 2,49957 3,50000 4,34360 4,49790 4,53220 ; data want; set have; by count_cluster; _prev_salary = lag1(max_salary); if first.count_cluster = 0 then do; Max_Salary_previous_grade = _prev_salary; end; drop _prev_salary; run; Bruno ... View more

Good afternoon and thank you for your responses: This is a partial dataset and the code: I have three people, 63, 64, 65 years old. ods html; data employees; input age @@; datalines; 62 63 64 run; I want to know what percentage of my workforce will retire in 1 year and in 2 years from now. the variable percentage at delete_dummy=0 should give me the answer.   data retirement; set employees end=end; do year=1 to 2; if age<(65-year) then numerator_factor=1; else numerator_factor=0; denominator_factor=1; denominator+denominator_factor; numerator+numerator_factor; Percentage=1-(Numerator/Denominator); delete_dummy=1; output; if end then do; delete_dummy=0; percentage=percentage; output; end; format percentage percent8.2; end; run; proc print data=retirement; run; ods html close; This is what I get Obs age year numerator_factor denominator_factor denominator numerator Percentage delete_dummy 1 62 1 1 1 1 1 0.00% 1 2 62 2 1 1 2 2 0.00% 1 3 63 1 1 1 3 3 0.00% 1 4 63 2 0 1 4 3 25.00% 1 5 64 1 0 1 5 3 40.00% 1 6 64 1 0 1 5 3 40.00% 0 7 64 2 0 1 6 3 50.00% 1 8 64 2 0 1 6 3 50.00% 0 This is what I would like to get Obs age year numerator_factor denominator_factor denominator numerator Percentage delete_dummy 1 62 1 1 1 1 1 0.00% 1 2 63 1 1 1 2 2 0.00% 1 3 64 1 0 1 3 2 33.33% 1 4 64 1 0 1 3 2 33.33% 0 5 62 2 1 1 1 1 0.00% 1 6 63 2 0 1 2 1 50.00% 1 7 64 2 0 1 3 1 66.66% 1 8 64 2 0 1 3 1 66.66% 0 Any ideas? Thank you! ... View more