The SAS Output Delivery System and reporting techniques

SAS graphic statements storage

Super Contributor
Posts: 260

SAS graphic statements storage

Hi all.
Since this is a SAS/Graph question, I'm not sure that it is the correct place to post it. Anyway, since it will be read be a lot of SAS experts, I think it's worth a try, isn't it ?

When you define some global graphic statement like AXIS1 or SYMBOL3, the definition you provided is stored in the WORK.SASGOPT catalog.
Is there a way to store it elsewhere, as you can do with macros or formats catalogs ?

Thanks in advance.
Super Contributor
Posts: 297

Re: SAS graphic statements storage

You could store it as a "program" in a directory somewhere and then call it in via a %Include statement as needed ...
Super Contributor
Posts: 260

Re: SAS graphic statements storage

Hi Bill.

Thanks for that idea, but I was wondering if there was some set of options like the FMTSEARCH and SASMSTORE options to give SAS a library where such an defintions catalog would be store.

I do agree that there are numerous ways to record definitions otherwise that try to save and restore such a catalog. Let's say that my question is rather of the theorical kind, than the "I have to solve this problem at any cost" kind.
Not applicable
Posts: 0

Re: SAS graphic statements storage

There may be Olivier, and some experimenting is called for.

I will tell you that by default the library GDevice0 is not assigned, but when you create your own device drivers for SAS/Graph, you then assign that library, copy SAS supplied drivers over to that library and modify them, and they then take precedence over the Device entries in SASHELP.

Never having played with saving AXIS and similar statements, I can't say with certainty that there is a catalogue search sequence you can use as a solution, but if there isn't then I'm willing to bet that one can craft a work around. One can certainly write two macros to save and recall definitions between the WORK catalog and a permanent version. I shall look at that, although since such definitions as AXIS are specific to particular outputs, I have some doubts about the need for the solution in my work.

Kind regards

Ask a Question
Discussion stats
  • 3 replies
  • 3 in conversation