@SuryaKiran ,   Wonderful! Thanks a lot!   Thank you.   - Dr. Abhijeet Safai   ... View more

See a gchart solution in rakul's thread. Are you two working together, or just happen to be wanting the same graph?    https://communities.sas.com/t5/Graphics-Programming/Need-to-change-labels-in-proc-gchart-vbar/td-p/674945 ... View more

Please post your question in a new thread. As you have noticed, nobody saw your post and could therefore answer to it; I only stumbled across it by accident. Please do not hijack question threads that are not yours and are hopelessly out of date. ... View more

There are two possible solutions to this. What do you want if you have a string like 'a==b=c'?   Do you want the result to be '=b=c' (which is not a SCAN solution, as consecutive delimiters are treated as one by SCAN), then the INDEXC solution suggested by @novinosrin is the way to go, if you want the result 'b=c' then take a look at the CALL SCAN routine, as suggested by @r_behata. ... View more

You need to provide a lot more details of your actual problem. Please post a small sample data set and expected output, which would help a lot more. For example, how do you know what values should exist? To me if I find Jan, Feb, March, I would say March is not correct because its more than 3 letters? How do you know you need 4 rows? ... View more

Having the expected outcome is a big help.  Here's what I recommend:   proc sort data=have; by TN; run;   data want; do until (last.tn);    set have;    by tn;    if EII=0 and ND=0 then both_zero='Y';    else if EII=1 and ND=0 then eii_only = 'Y'; end; do until (last.tn);    set have;    by tn;    if EII = 1 and ND=1 and (eii_only='Y' or both_zero='Y') then output; end; drop both_zero eii_only; run; ... View more

data have; input ID var1 $ var2 ; cards; 111 aaa 100 111 aaa 101 111 . 102 111 . . 222 bbb 101 222 bbb 102 222 . 103 ; data want; set have; length new_var1 $ 80; retain new_var1 new_var2; by id; if first.id then call missing(new_var1,new_var2); if not missing(var1) then new_var1=var1; if not missing(var2) then new_var2=var2; if cmiss(new_var1,new_var2)=2 then call missing(new_var1,new_var2); run; ... View more

And here is my Log report of  my test with 30 Million observations   932 /*Compress char and keep digit*/ 933 data want; 934 set have; 935 val_1 = input(compress(Value,,'kd'),best.); 936 run; NOTE: There were 30000000 observations read from the data set WORK.HAVE. NOTE: The data set WORK.WANT has 30000000 observations and 3 variables. NOTE: DATA statement used (Total process time): real time 6.14 seconds cpu time 6.14 seconds 937 /*extract num from currenct and apply informat */ 938 data want3; 939 set have; 940 val_1 = input(substr(value,2),8.); 941 run; NOTE: There were 30000000 observations read from the data set WORK.HAVE. NOTE: The data set WORK.WANT3 has 30000000 observations and 3 variables. NOTE: DATA statement used (Total process time): real time 2.85 seconds cpu time 2.73 seconds 942 /*comma informat */ 943 data want1; 944 set have; 945 val_1 = input(Value,comma10.); 946 run; NOTE: There were 30000000 observations read from the data set WORK.HAVE. NOTE: The data set WORK.WANT1 has 30000000 observations and 3 variables. NOTE: DATA statement used (Total process time): real time 3.67 seconds cpu time 3.57 seconds   ... View more

Hi Allaluiah, I tried to solve the problem in other way. Thank you for your comment. Here is the simple code .. proc sql; create table want as select subject_id ,visit_day ,expected_visit_date,actual_visit_date , case when expected_visit_date <= max(actual_visit_date) then coalesce(actual_visit_date,expected_visit_date) else . end as new_date format=mmddyy. from have group by subject_id order by subject_id,visit_day; quit; ... View more