I think it depend on what version of MC is installed. I don't remember any more wich one we had, anyway, we upgraded it and the export of G2 report has began to work. ... View more

libname abc slibref=abc server = user ; You need the slibref statment 😉 ... View more

Yuo have start as numeric. call symput('start',input("&st",mmddyy10.)); Right? If so don't use the input function. It trasform char to alphanumeric. It's hard to expain using a formum as a chat ... View more

Thank you David. I've checked that I miss some module. I hope that as soon as possible they'll be installed. I don't manage the SAS Depot. ... View more

I miss something. regis_dt and stop_dt are SAS date? If so they are numeric and their value is the number of day sinc january 1st 1960 You can check changing the format to best. You should trasform the '20140301' to a date in this way using the input function input('20140301',yymmdd8.) So the datastep should be before the proc sql. One of most difficult thing to understand in SAS is how it handles the date. It's very powerfull one you'll understand it ... View more

trasform the variable char with the '1' '2' .. '100' in numeric and appaly the formmat zn. (n it,2 the max number og digit, see the format z. on the documentation) so that it's shown as '001' '002' ... '100' ... View more

Very dirty. Create a new variable with the ritgh order. order by it. In the wrs not order but with the order given by the infomap. or use tha variable just for the order. If you do for just one varaible could be onest Another way cam be a format with the righ order and order by format on the infomatp but showing no format at all on the WRS. You should ask if is possible to order in a 'custom way' and I dont't if is possible. ... View more

You can use an annotate dataset. There's some documentation on the PROC GMAP. ... View more

It's an append. Ypu can do with a datastep data new;   set pi_1 P_new; run; Or with the proc append. Proc append base=P_1 data=P_new force; run; The code is written by memory so some error could be in it. Test it and change it. ... View more

I'll use the retain and the first and last. Something like this (I don't know if works because have written it by memory, but you can start by it) proc sorta data=; by MemberID program_start_date;run; data output;   set input; by MemberID;   retain xxx  _sum  .;   if first MemberID then do;     xxx=program_end_date;   end;   if program_start_date=xxx+1 then do;     _sum=sum(_sum,duration);     xxx=program_end_date;   end;   if last.MemberID then output; run; The problem is if yuo need the max totalday or the last total_day. My code do the last, but with some changment you can do the first option ... View more

Hem ... what's the code? Proc format? ... View more

Or the exist function to check the the existence. %sysfunc(exist(<data_set>)); If you use the open(), as suggested by LinusH, it's better to use the close after the check, otherwise the DataSet could be locked. ... View more

May be I've havente understund your problem but if the istuation is this a 100  10 b   50  30 c  20   40 d 10    20 Then sum c = 100*10+50*30+ 20*40+10*20 In this case not use the aggregate measure but the 'normal measure sum c = (100+50+20+10)*(10+30+40+20) In this case use the aggretagate measure with sun for group ... View more

The answer is easy: no. You can rename the two column in input data final;   merge Directa Folios (rename=(Burks=folios_Burks Blartd=folios_Blartd));   by xxx; run; If you append you can. ... View more