Thank you so very much for this response. ... View more

Tom, I used this code:  data want; do until(last.id); set once; by id; length first_site $8; first_site=coalescec(first_site,site1); end; do until(last.id); set once; by id; output; end; run;   I was wondering how I could incorporate a county code for the patient to get the same output by county?  The table output would look like this:  patient Inn342 visited two counties with most recent in 01 is Sha and in county 19 is Dha.     updated code to include county and blanks. (thank you) data once; input ID $ county $ visit1 $ site1 $ ; CARDS; Inn342 01 . . Inn342 01 Anc Sha Inn342 01 Drr Loc Pan322 19 . . Pan322 19 Cli Dha Pan322 19 Omn Sha ; run; ... View more

Thank you so much KSharp!! I so appreciate your help. This was exactly what I was looking for. thank you again! ... View more

I'm not sure what you mean by the second question, but you can use PROC MEANS or TABULATE and save the output to a data set instead of having it displayed if that's what you're asking about. Use the OUT= option on PROC TABULATE or there are a variety of options with PROC MEANS. proc tabulate data=have out=AGG_SUMMARY; ... View more

For example: data have; input ID $ Age Quad $; datalines; Smi12 3 N Cha890 2 N Tse2987 18 S Bra7650 22 N Gon12409 31 N Rod23433 35 S ; proc format; value multiAge (multilabel notsorted) 0-999 = [7.0] 0-5 = " 0-5" 6-18 = " 6-18" 19-35 = " 19-35"; run; proc summary data=have nway; format age multiAge.; class age / mlf order=formatted; class quad; var age; output out=want n=freq; run; proc print noobs data=want; var age quad freq; run;   ... View more

@SannaSanna    Let's discuss the first problem you identify:   @SannaSanna wrote: ... stuff deleted ... but I have also used Proc SQL to tie the tables together and still missing records. "Missing records" is way too amibiguous to diagnose.  But from what you have shown us, any missing records are not due to size of the data sets.  Please provide an explicit example of a record you expected in the result, but was missing.  And provide the initial records from which you expected it to be produced.    You should be able to extract those records into a sample, and then quickly run both your SQL code and DATA step merge code that resulted in missing records.    Show us those codes, and the log file that went with them.   Then diagnosis (and maybe prescription) is possible.     ... View more

Hi-   I think I almost got it to run correctly.  The code failed untimately.   The comparison code-  why is my data being converted from Character to Numeric?    The data contents within TableA and TableB are all string/characters/YN and no numeric.  And does the data need to be sorted first in the merge step?  << got error message too   ... View more

This is really out of my league. I would guess that there wasn't enough disk space. I suggest you post a new topic with the full query and the error message. ... View more