Pass the log state to the histogram machinery

[cphyc]

This fixes #1962.

[cphyc]

Using the same dataset as #1962 I know have the following plot

[astrofrog]

Thanks for the PR! Just a small comment below, and otherwise can you add an entry under 0.14.3 in CHANGES.md?

[astrofrog]

This might give strange results if both xmin and xmax are zero for some reason - I think it might be best to actually just return an array of NaN values if either xmin or xmax are negative rather than modify the limits. You can construct this array efficiently with:

broadcast_to(np.nan, bins)

could you update this here and below?

[codecov]

Codecov Report

Merging #1963 into master will increase coverage by 0.05%.
The diff coverage is 50%.

@@            Coverage Diff             @@
##           master    #1963      +/-   ##
==========================================
+ Coverage   87.48%   87.54%   +0.05%     
==========================================
  Files         244      244              
  Lines       21716    21775      +59     
==========================================
+ Hits        18999    19062      +63     
+ Misses       2717     2713       -4

Impacted Files	Coverage Δ
glue/viewers/scatter/state.py	91.82% <ø> (ø)	⬆️
glue/core/data.py	90.22% <50%> (ø)	⬆️
glue/viewers/matplotlib/layer_artist.py	95% <0%> (-2.5%)	⬇️
glue/viewers/histogram/layer_artist.py	93.05% <0%> (-0.7%)	⬇️
glue/_mpl_backend.py	100% <0%> (ø)	⬆️
glue/_deps.py	66.25% <0%> (+2.45%)	⬆️
glue/viewers/matplotlib/viewer.py	97.82% <0%> (+2.55%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9ab4614...4d17185. Read the comment docs.

[astrofrog]

I'll go ahead and merge this, thanks @cphyc!