scikit-hep
Actual tests
Would be good to have some tests that can actually fail instead of depending on visual checks.
We could check for an exact match in an output string of a histogram, but I'm not sure that is very useful. I guess it would add more extra work (to keep the comparison string up-to-date) than it's worth.
That's fair enough. One idea might be to profit from the fact that you interoperate well with boost-histogram and simply add a couple of comparisons after the conversion. It's not a pure/direct test, but that's probably OK enough. Dunno if @henryiii would have some other idea ... Otherwise you can probably close this task as done, I guess.