Adam Gutglick
Adam Gutglick
When evaluating expressions in `vortex-datafusion`, if we don't have specialized implementation we'll have to canonicalize arrays over and over which is very wasteful.
My 2c here - I wished I could fork Criterion multiple times, and it seems like Divan is actually maintained which is a big plus in my book
The first thing I would try is to create the runtime as part of the setup, and then call `Runtime::block_on` or it's equivalent in the actual benchmark code. From reading...
I think the difference here is big vs small benchmarks, where I would love for smaller benchmarks to be divan/criterion based, and the larger ones can be based on our...
Is there anything to this issue besides changing the default `datafusion.execution.collect_statistics`, fixing any tests that rely on the default value being `false` and and changing docs accordingly? Would be glad...
Definitely, I'll run [our benchmarks](https://github.com/vortex-data/vortex/pull/3560) once I get all tests passing here.
I'm getting a lot of sqllogictest failures, is there a reason to think there something weird going on? I was somewhat open to the idea its all fine until I...
Added a short upgrade note
Got a similar test failure to #16448 (issue filed in #16452). I have to conclude its personal at this point, I'll try and find some time to dig into it.
Our benchmarks show this change fixes the performance regression we saw - https://github.com/vortex-data/vortex/pull/3567