OscarTHZhang commented on issue #94:
URL:
https://github.com/apache/arrow-datafusion/issues/94#issuecomment-1207661351
Hi @alamb,
Here are some questions come up my mind
- At what granularity a benchmark should be operating at?
- For aggregate, for example, do we also need to implement the micro
benchmarks for all the aggregate functions and also all the physical aggregate
expressions (like correlation)?
I think we can divide the micro-bench into 2 types (as described above)
- Single Operator bench
- Targeted SQL
For all the aggregations, if we are going to implement them all, we can
simply write the targeted SQL benchmarks.
For operators that operates on column- and table-granularity with output
sill as columns and tables, set up the single operator bench for them, such as
merge, join, filter.
How does this sound? Anything missing?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
