Imrprove label plotting code #271
Labels
feature
New feature or request
good first issue
Good for newcomers
infrastructure
Non-quantum things to improve the robustness of our package, e.g. CI/CD
Is your feature request related to a problem? Please describe.

The labels in our benchmarking plots, which show the compiler name and version, can get very messy:
and our current plotting code has a lot of fiddly adjustments involved, which often fail to converge.
Describe the solution you'd like
Modify the code so that the labels successfully plot without overlapping. Because the current code is pretty unwieldy, it may help to break it out into a separate module outside of ucc/benchmarks/scripts/common.py.
We should probably test out this package, which is meant to automate text placement to minimize overlaps: https://github.com/Phlya/adjustText.
Describe alternatives you've considered
There may be other tools or packages which could do this, but I have mainly focused on options involving matplotlib annotations.
Additional context
Our benchmark plots are an important part of transparency for our project, and making them clear and legible is a necessity.
The text was updated successfully, but these errors were encountered: