Hi @novinosrin, Neat. I have a very rudimentary understanding of it but see the endless possibilities in the ways it is used. I am studying and perhaps one of these days will be able to post a solution with it. Fascinated and yes, I am looking into the book by @hashman and @DonH. ... View more

Hi @A_Halps, There is a neat trick that I learnt just a few days ago from @novinosrin. Go ahead and give this a try. I have added a satellite variable WEIGHT to the HAVE dataset and tested. A partial print of the final data is produced.   data have; input Study_ID $ Year weight ; cards; 001 9 50 001 9 50 001 10 51 001 12 52 001 15 54 002 2 60 002 6 62 002 8 64 ; run ; proc sql ; create table temp as select study_id, min(year) as minyear from have group by study_id order by study_id ; quit ; %let maxyear = 15 ; *set this value to what you desire ; data template ; set temp ; by study_id ; do year = minyear to &maxyear ; output ; end ; run ; data want ; merge have template ; by study_id year ; drop minyear ; run ; data real_want ; update want (obs=0) want ; by study_id ; output ; run ; Obs Study_ID Year weight12345678 001 9 50 001 9 50 001 10 51 001 11 51 001 12 52 001 13 52 001 14 52 001 15 54 ... View more

Hi @deengyn, This is helpful but need more answers. As a preliminary step, you can take my program for the first patient and add 1 or 2 more patients with the data that you currently HAVE. Next, let us also know what you would like to see as the end result of that dataset with DIFF generated. Let us keep this to convention and call it the WANT dataset. From what you have clarified in (2) above, I see that you may need more than one DIFF value. DIFF1 - from General Visit. DIFF2 from Initial Surgery. Finally, from (3) would you be able to add a sample patient to cover that scenario and how it would affect DIFF(s) values. You would also need to generate the WANT dataset with code similar to the one I provided for generating the HAVE dataset above. One last piece. The dataset (picture) from your first post does NOT represent the code that you provided later. If that is the result of the sort below then for Patient 51, the first SERV_DATE value would be 25SEP2017. *Formatting; proc sort data = plbr out = plb_sorted; by pid serv_date rec_num ; run; ... View more

For 001 the first two observations are duplicates. Is that how it appears in your real data or a typo? ... View more

Can really use code for the input dataset. I have created for the first patient. A couple of questions.   1. For DIFF, are both dates included - most recent general visit and first plastic procedure? You will need to add a day to it.  2. Once a procedure is done, do you care about DIFF value for a later procedure for the same patient? 3. Could there be a general visit between procedures? If the dataset is pre-sorted by PID and DATE then the solution could be simpler.   The above are unknown so the solution is based on some assumptions on my part. data have ; infile datalines missover ; length pid 8 serv_date 8 summary_cod $3 pl_cod $12 ; format serv_date date9. ; input pid serv_date:date9. summary_cod $ pl_cod $ ; datalines ; 51 26SEP2017 BR . 51 25SEP2017 SBR . 51 19MAR2018 . removal 51 19MAR2018 . mastopexy ; run ; data want ; do until (last.pid) ; set have ; by pid ; format recent_date date9. ; if not missing(summary_cod) then recent_date = max(serv_date,recent_date) ; if not missing(pl_cod) and nmiss(diff,done)=2 then do ; diff = serv_date-recent_date ; done = 1 ; end ; *else diff = . ; output ; end ; run ; ... View more

Oops. Looks like Cynthia has already posted a reply. I tried reloading the page to see if it was already answered but I didn't see any. Lo and behold when I posted the other replies appeared.  ... View more

I ran your code and I do see a TOC collapsed in the RTF. It is just not as apparent as in the Results window where you clearly see a downward arrow. You have to right click by placing the cursor BELOW the Table of Contents area and choose Update field. Not a very desirable behavior but it is expected. For more on this see: https://documentation.sas.com/?docsetId=odsug&docsetTarget=p1vvsv8ucnjzjnn1wq5wrlp74mdb.htm&docsetVersion=9.4&locale=en#p0pg2tmkd9x2own1rsrokb83vcsy   See Example 1 and scroll a bit down to the end of the example. ... View more

Many possible solutions. Here is one with the now popular DOW loop.    data have ; infile datalines missover ; length study_id $5 year samh demp 8 ; input Study_id $ year samh demp ; datalines ; 001 5 2 . 001 6 3 . 001 9 3 . 002 7 2 . 002 8 4 . 002 10 3 . 002 16 3 . 003 20 . 4 003 20 . 5 003 20 . 4 ; run ; data want ; do until (last.study_id) ; set have ; by study_id ; if nmiss(samh,samh_adj)<2 then samh_adj = max(samh_adj,samh) ; if nmiss(demp,demp_adj)<2 then demp_adj = max(demp_adj,demp) ; final = sum(samh_adj,demp_adj) ; output ; end ; run ;   ... View more

Your problem is perhaps less complicated and could be solved by a procedure like TRANSPOSE. It is not clear from your description. However, if you really want a solution as described, here is one solution.   data have ; input var1 var2 var3 ; cards ; 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 ; run ; data want (drop=_:) ; set have ; retain _v1 - _v3 ; array _v(3) ; array var(6) ; if mod(_n_,2)=0 then do ; do _i=4 to dim(var) ; var(_i) = var(_i-3) ; var(_i-3) = _v(_i-3) ; end ; output ; end ; do _i = 1 to dim(_v) ; _v(_i) = var(_i) ; end ; run ; ... View more

You may want to look at Rick Wilkin's example of how he has used SAS Jedi's macros to keep variables with a certain suffix. Looks like what you want.  https://blogs.sas.com/content/iml/2019/10/28/regular-expressions-variable-names-sas.html   You will need to download the macros from Jedi's post and they may be useful later for a lot of other tasks. https://blogs.sas.com/content/sastraining/2017/08/29/sas-variable-lists-by-pattern/ ... View more