If anyone wants to have a go at this, #8993 and #9889 show what was done to ax.hist to make this change.

Hi @dstansby, I would like to do this. This PR is for both replacing and depreciating, right?

I think you need to deprecate the normed kwarg, and add a density kwarg and make it compatible w/ numpy. Don't remove the normed kwarg... Not 100% sure if normed and density are meant to do exactly the same thing...

@dstansby @jklymak I noticed that np.histogram function of numpy has newer kwargs with depreciated normed argument but np.histogram2d (here) still works on normed and not density. How do we replace the arguments for ax.hist2d then?

I’d argue for consistency with numpy. If they don’t used notmed I guess we shouldn’t.

Agree with @jklymak (maybe I should convince numpy should change for API consistency 😉 )

numpy 1.15 introduced a 'density' kwarg to hist2d (numpy/numpy#11529). Even if we don't require numpy 1.15, we can always just introduce the 'density' kwarg and pass it as 'normed' to numpy.

Hi All,
I just submitted the pull request for this issue here. This is my first request and please let me know if something is missing.

Closed by #13128.