There are already several answers in the thread you started Monday: https://communities.sas.com/t5/SAS-Enterprise-Guide/SAS-Enterprise-Guide-multiple-values-in-Prompt/m-p/974391#M46156   Please review, and respond in that discussion thread if you have followup questions. ... View more

I think this could be done by using a discrete attribute map, to map each group value to a color.   data _test; do grp=1 to 6; do ind=1 to 100; yvalue=sin(ind/10)+grp; gx=grp; output; end; end; run;quit; data mymap ; input value linecolor $10. ; retain id "grp" ; cards ; 1 Red 2 Blue 3 Yellow 4 Green 5 Purple 6 Black ; ods layout gridded columns=3 rows=2 advance=bygroup; proc sgplot data=_test noautolegend uniform=xscale dattrmap=mymap; by grp; series x=ind y=yvalue / group=grp lineattrs=( thickness=2 pattern=solid) attrid=grp; refline 30/axis=x lineattrs=(color=red thickness=2 pattern=solid); run; ods layout end;   ... View more

As others have explained, when you use a multi-valued prompt, the values are stored in multiple macro variables.   This paper by Joe Flynn has a section explaining an approach: http://support.sas.com/resources/papers/proceedings11/050-2011.pdf   Based on that approach, I created a function-style macro which returns a delimited list of the selected items:   %macro ConcatenateSelectionList (prompt ,dlm=%str( ) ); /*============================================================================== Macro: ConcatenateSelectionList Abstract: Macro function that returns a delimited list of items chosen in a check all that apply stored process prompt. Description: Stored processes pass CHECK All That Apply parameters in an odd fashion. If parameter is named PROMPT then &PROMPT_Count always exists, holds number of options selected (0-n) &PROMPT1 - &PROMPTN are created if count>=2, if count=1 or 0, they not created &PROMPT is created if count >=1, if count=0 it is not created. Usage: %let sitelist=%ConcatenateSelectionList(site) ================================================================================*/ %local i return ; %if &&&prompt._Count >= 2 %then %do i = 1 %to &&&prompt._Count ; %if &i=1 %then %let return=&&&prompt&i ; %else %let return=&return&dlm&&&prompt&i ; %end ; %else %if &&&prompt._Count = 1 %then %do ; %let return=&&&prompt ; %end ; %else %if &&&prompt._Count = 0 %then %do ; %let return= ; %*null ; %end ; %else %put ER%str()ROR: USER &prompt._Count=&&&prompt._Count ; &return %mend ConcatenateSelectionList;   ... View more

What is your concern with using NLOBSF?   Years ago I was inspired by this paper by Jack Hamilton: https://support.sas.com/resources/papers/proceedings/proceedings/sugi26/p095-26.pdf   And wrote a obscount macro, the core of which is to use NLOBS if SAS knows the number of records and there is no WHERE options, else use NLOBSF: %if %sysfunc(attrn(&idnum, ANOBS)) and %sysfunc(attrn(&idnum, WHSTMT))=0 %then %do; %*If sas knows number of obs, and there is no where statement, get nlobs; %let nlobs = %sysfunc(attrn(&idnum,nlobs)); %*get number of obs; %end; %else %if %sysfunc(attrn(&idnum, ANOBS))=0 or (1<=%sysfunc(attrn(&idnum, WHSTMT)) and %sysfunc(attrn(&idnum, WHSTMT)) <=3) %then %do; %*If sas doesnt know number of obs, or there is a where statement, get nlobsf (this iterates through dataset); %let nlobs = %sysfunc(attrn(&idnum,nlobsf)); %*get number of obs; %end; Yes, DOSUBL is slow compared to most other functions, because it's doing a LOT of work (creating and destroying the side session to run execute the code).  ... View more

Can you replicate the problem with the little step you showed?   Something like:   data want ; set temp_01 ; age_on_date = int(yrdif(dob, date_01, 'AGE')) ; if missing(age_on_date) then put dob= date_01= ; format dob date_01 ; run; If so, then please post the full log from running that step.  Well maybe, not the full log if it's 1000s of lines from the PUT statement.... ... View more

Check if your log says "Note: missing values were calculated at... "  That would suggest that either DOB or Date_01 are missing values for some records.  Many people treat that missing values note as an error message. ... View more

I'm not familiar with those macros, but I see they are provided by SAS.  I would probably start by running the code with the system option MPRINT turned on, then review the proc template code to look for the yaxisopts option and see if the label is there.   Alternatively,  for custom graphs I often find it easier to have the PROC output the data, and then write my own SGPLOT or GTL to generate the graph. ... View more

Looks like you can pull this from dictionary.tables:   proc sql ; select memname, sortname, sorttype, sortchar from dictionary.tables where libname="WORK" ; quit ; See also this thread, which uses dictionary.columns :  https://communities.sas.com/t5/SAS-Programming/Keeping-the-defined-sortedby-values-in-the-metadata/td-p/823697 ... View more

Make sense.  Overwriting datasets: data foo; set foo; run; can cause all kinds of confusion.  Glad you got it sorted. ... View more

One simple check is to see if a WHERE statement will find records in both datasets. If you run: proc print data=person_file; where person_contract='90091'; run; proc print data=contract_file; where contract_id='90091'; run; Do both PROC PRINT steps return rows?  Hopefully one of them does not, meaning the value '90091' is not in one of the datasets.  ... View more

It's often helpful to run PROC FREQ to check the values of your variables.   If you run: prof freq data=work.tsa_nodupkey; tables Incident_Date*Date_Received/missing list; format Incident_Date Date_Received year.; run; are the results informative? ... View more

Check the header of the file.  By default the title is not written to a cell in the spreadsheet, instead it is written to the header.   I think in Excel you can still do Page Setup to see the headers. ... View more

I assume that's just the beginning of the email?  So basically the body of the email is showing you the HTML code instead of rendering the HTML?   Since you have both M6 and M8 available, you might try writing the MSOFFICE2K output to a file, and then comparing the files you get from M6 and M8 to look for differences.  I'd be surprised if SAS intentionally changed the MSOFFICE2K template.   Once you have the differences, you could send the sample code and output files to tech support.     ... View more