Henry Schreiner
Henry Schreiner
That's mildly annoying but not terrible. Though I'm guessing there's a way to avoid it: https://github.com/pypa/build/pull/443 for example.
I very much thing we should do the embeddable PNGs, it just hasn't been done yet. :) The idea of the repr is a simple, zero dependency display. Having docs...
So these return histograms, rather than values?
See https://github.com/scikit-hep/scikit-hep-tutorials/blob/4697144a3fa1e7a1c41824dbe3d9a1e9ecac89c3/content/histogram/UprootExample.ipynb
Also, stack.name should return and allow the setting of the names.
Raw setting is the only thing left. `s[0] = h`, where `h` gets checked to see if it is valid.
This is done (at least for the things listed above), but we should add the new stuff to the readme.
The repr is normalized, so it's leaking NaN's in, which breaks it - we should filter/mask those. Will have to investigate the integer division one!
I think a structure like this: ```python [ [{'Jet_pt': 1, 'Jet_eta': 2}], [{'Jet_pt': 1, 'Jet_eta': 2}, {'Jet_pt': 1, 'Jet_eta': 2}], ] ``` Converts to this hist fill: `Jet_pt=ak.flatten(awkarr['Jet_pt']), Jet_eta=ak.flatten(awkarr['Jet_eta'])`
`h2[...] = h1.view(flow=True)` allows assignment. Histogram histogram assignment would probably need some checking on the axes which is not implemented yet.