Benchmarks for Scikit-HEP

[pfackeldey]

I attempted setting up a suite of benchmarks recently for Awkward Array: scikit-hep/awkward#3548, but I'd like to think about a way to expand this to other packages/tools from our ecosystem as well.

We also have https://github.com/scikit-hep/integration-tests which is a repo to run test suite with the current main branches of various scikit-hep packages against each other to make sure there are no problems with new releases.

I'd be interested to think about setting up benchmarks for regression testing for our ecosystem. In addition it's nice to track performance improvements over several releases - there's potential for nice visualizations here!

Tagging @lgray as you already raised this idea here: scikit-hep/integration-tests#4

[eduardo-rodrigues]

Hi. +1 on the initiative! It would be great to expand on what we have in integration-test, which for now only host a little test running on new PRs whereas we would want to have nightly tests or test run on a regular basis.