Parallel Processing in SAS

[Aexor]

Few Question in optimization :-

1. how can we achieve Parallel Processing in SAS.?

2. How can be solve space issue after executing a program in sas 

3. how can we optimize a process involves in reading sas data set ?

 Already tried compress, keep,drop, data _null_ 

4. Latency in SAS.

these are few questions i came across, i am looking for any suggestion or documentation which talks about these topics.

Thanks!

[Kurt_Bremser]

Enterprise Guide allows the use of parallel SAS sessions within a project. And a lot of SAS procedures support multithreading out of the box, especially SQL

On the storage side, using SPDE libraries can speed up I/O massively.

Space issues are dependent on the data you have to use. Some rules of thumb:

• Reduce data early. Only keep variables and observations that are needed later.
• Summarize data early, where appropriate.
• Use compression (selectively).
• Remove intermediate tables when they are no longer needed.

To optimize a process, the good doctor has to get to know it. So we need to see a code piece that eats lots of time and space to make suggestions for improvement. Also do not forget to add information about the data involved (variable sizes, observation numbers).

Define "latency". Latencies are found in network connections, storage, and a gazillion of other things.

[Aexor]

Thanks much!! It's very helpful

[yabwon]

About: "1. how can we achieve Parallel Processing in SAS.?"

If you have XCMD turned-on on your SAS you can check out my little tutorial about running code in parallel in SAS:

https://pages.mini.pw.edu.pl/~jablonskib/SASpublic/Parallel-processing-in-BASE-SAS.sas

If you have any questions on that material let me know.

All the best

Bart

[Aexor]

Thanks much!

[SASKiwi]

Parallel processing is also possible using SAS/CONNECT. Refer to the SAS/CONNECT documentation for more details.