Issues on -runby- command CPU using

[Evelyn Zou]

Hi everyone:
When using 24-core version of Stata 17, I notice that the -runby- command uses only 4% of CPU (which means it only uses one core), however, other commands can use up to 50% CPU. At the same time my dataset is very large, so the code runs very slowly. Is there a way to set the -runby- command customized to multi-core?
Any help would be really appreciated!
Best,
Evelyn Zou

[Nick Cox]

I think the answer is No. runby (from SSC) doesn't have such an option beyond what is implied by the option allocate(). IIRC, there is no way that community-contributed code can have such an option.

[Evelyn Zou]

I think the answer is No. runby (from SSC) doesn't have such an option beyond what is implied by the option allocate(). IIRC, there is no way that community-contributed code can have such an option.

Thank you for your warm reply! I finally worked it out, by rewriting the code in a no-loop way.
I guess maybe the problem is not about the -runby- command, but the commands I written in the -runby-. It may be a very bad coding style that loops each sample.