One of my new work friends says this frequently. But the statement is lost in his translation from Portuguese, it is not that he does not believe me, it is just that he has questions about how things happen so he can learn. I debate on which version of this statement I get from others, especially when it comes to the "magic" of using a SAS Stored Process within a SAS Information Map. In summary, I get two main questions from the tip/example I posted back in '09 about using Stored Processes to use SQL Pass-thru from Information Maps:
- The first concern is with concurrancy
What happens when multiple people hit the SAS Information Map at the same time? In the example, the most important statement is:
libname BIOUT (work);
This reassigns the library BIOUT to the location of work. If you are not familiar with SAS, this is the individual location for each instance of SAS that is distinct, temporary, and is completely removed if the SAS session ends normally. Therefore, reassigning your BIOUT location to (work) ensures that the table created dynamically by the SAS Stored Process is the file only for your run.
- Next is the question that no physical file exists
Again with the use of libname BIOUT (work); everything is cleared when the session closes. So a physical file exists, but only briefly.
Now that you have no doubts about using this technique, how do you plan to leverage the power of SAS Stored Processes? If you need help getting started, I previously posted four ideas on how SAS Information Maps can leverage Stored Processes "Using Stored Processes to Improve Information Maps".