PR for v.1.9.4

[hcpchris]

• Closes Jupyter Error on rendering chart from_pandas when df has NaN values #200 . Incorporated bugfix for handling np.isnan() operations that generate a TypeError when casting in safe-mode (the default). Thanks to @ThomasGl !

[codecov]

Codecov Report

Attention: Patch coverage is 66.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 87.95%. Comparing base (b6925a1) to head (7b12281).
Report is 6 commits behind head on master.

Files with missing lines	Patch %	Lines
highcharts_core/js_literal_functions.py	66.66%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #203      +/-   ##
==========================================
- Coverage   87.96%   87.95%   -0.02%     
==========================================
  Files         210      210              
  Lines       31164    31166       +2     
  Branches     2552     2553       +1     
==========================================
- Hits        27413    27411       -2     
- Misses       2791     2793       +2     
- Partials      960      962       +2     

Flag	Coverage Δ
	87.95% <66.66%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[ThomasGl]

Hey Chris,

Thank you so much for all your help in finishing up this PR! Also for merging it and editing the contributors to the project list. Prehaps there was something missing there, as I am not currently listed on the contributers section of the repo, not sure if I have to do some action on my side.

[ThomasGl]

Also I will check the fix on the safe casting type you made. I don't recall ever managing those on python before

[hcpchris]

@ThomasGl : Hmm - odd. It seems that Github is listing you as a contributor on the release (see here ) but for some reason not listing you on the overall Github contributors list. That seems decidedly strange to me.

In any case, I've included you on the contributor list in the docs which get listed in a couple places there.

And as for the safe casting change, basically there were two types of value that were getting handled improperly by np.isnan(), specifically Decimal types and None which I was quite surprised to learn that np.isnan() errors out on when in "safe" casting mode. You can see the checking that takes place in the unit tests that I added: one test basically tries all of the scenarios that were present in the object that was erroring out in the unit tests (used to determine which scenarios were causing the problem), and the second test actually tests that now they work as expected in get_js_literal().

[ThomasGl]

Hey Chris, I checked out the unit tests, they look great. It seems to me quite odd the safe casting type for the np.isnan(), I check on numpy lib if this behaviour is expected or some changes could be done in that matter.

As for the contributors mentions, I think there's some thing with how the PR was identified by guithub. I will try to contact them about it, in the meantime if you can check the git log and status of the repo there might be some indication there of what happend.

Thank you once more,
Thomas