Here is yet another paper https://support.sas.com/resources/papers/proceedings13/406-2013.pdf ... View more

@JMS  As it was mentioned before, without local SAS instance, your EG can only connect via SAS Metadata server. Getting the Spawner started is not enough.   Check these two, somewhat dated, papers to give you a good understanding on what's we have been trying to elaborate about the SAS Intelligence Platform.  https://support.sas.com/resources/papers/proceedings11/363-2011.pdf https://support.sas.com/resources/papers/proceedings/pdfs/sgf2008/051-2008.pdf    Hope this helps, Ahmed ... View more

Try this   LIBNAME exfiles XLSX '[drive:]\test_xlsx_lkup.xlsx'; options validvarname='any'; proc contents data=exfiles._all_; run; proc sql; select term ,'43986'n from exfiles.sheet1; quit; Additional examples can be found here Using LIBNAME XLSX to read and write Excel files    Accessing Excel files using LIBNAME XLSX   Hope this helps, Ahmed ... View more

Here is my approach - avoiding looping   data have(drop=str); input id$ dcode1 $ dcode2 $ dcode3 $; array codes (*) $ dcode1-dcode3; length str $32767; str=cats(of codes[*]); if (index(str,'J01XX05') gt 0) then do; J01_flag = 0; C03_flag = 0; end; else do; J01_flag = count(str,'J01')**0; C03_flag = count(str,'C03')**0; end; cards; 1 J01AA02 J01XX05 J01CA01 2 J01XX05 A10AB01 C03CA01 3 J01AA02 A10AB01 C03CA01 ; run; ... View more

For Huge data sets, I would avoid Proc Transpose. Using Data Step for fixed data structure can be faster and uses less data passes /* Transpose the data using Data Step */ data want(keep=varname levels y); length varname $7 levels $1; array vars (*) $1 var1-var3; set have; do i=1 to dim(vars); varname =vname(vars(i)); levels =vvalue(vars(i)); output; end; run; /* Summarize to produce your desired means */ proc summary data=want nway; class varname levels; var y; output out=want_sums(drop=_type_ Rename=(_freq_=freq)) mean=y; run; /* Remove unwanted data */ proc delete data=want; run; Hope this helps, Ahmed  ... View more

Thanks Chris, I'll keep that in mind for the future. Regards, Ahmed ... View more

In this case, it looks like your only option is using connection profile from within EG, similar to what @Patrick had illustrated with his attached screenshot. You'll need to contact your SAS Administrator and Request to be added/Registered in the SAS Metadata Server Request to be setup with the required authorizations to Folders, Data, Apps, and App Servers within the Metadata Server Get the connection details (Server Name, Port) Ahmed ... View more

@JMS  If and when you get the SAS/connect spawner running, then in EG you would do something like this in a program entry %let server = &p_linuxSrvr 7541; *<---- Replace the Port (7541) if different value being used by the Connect spawner; options comamid=tcp; signon server user="&p_user" password="&p_PASSWORD"; LIBNAME r_work slibref=work server=server; Rsubmit; /* SAS statement(s) */ EndRsubmit; /* When you are done */ signoff;  You'll need to replace &p_xxxx place holder variables with your actual values.   Hope this helps, Ahmed ... View more

Hi Patrick, Minor correction regarding your statement "The SAS Metadata server is part of the foundation package and you will always have it licensed. And it will always be up and running on a Server instance. That's what you need to connect to SAS even if using SAS/Connect (which allows to connect SAS/Servers not Client/Server)." That might be true for your software bundle, but it is not true at the organization I work for. My team is responsible for Installing and Configuring SAS software on Linux & Windows Servers, and the SAS Foundation bundle we install does not contain Metadata Server component! SAS Metadata Server is part of the SAS® 9.x Intelligence Platform, rather than SAS Foundation. I think it's all depends on how the SAS software bundle and license were contracted/forged between SAS Institute and the customer. Something to keep in mind, thanks, Ahmed ... View more

I did that, and still didn't work! ... View more