Closed milankl closed 2 months ago
@maximilian-gelbrecht the "automatic" testing is now simply by rerunning julia manual_benchmarking.jl
. It will recreate the benchmark/README.md
file so that we can regularly assess performance and compare across versions.
fixes #464
@vavrines this is a PR to address your point on performance documentation
Benchmark overview looks like
https://github.com/SpeedyWeather/SpeedyWeather.jl/blob/mk/benchmark/benchmark/
@maximilian-gelbrecht I found a manual benchmark that we can update every now and then easier to implement. Another option would be to let the documentation do this, in the end it's not that costly, most simulations run for O(1s) so it could be part of the documentation as well.