add bins argument to Histogram function

[zachcp]

The primary argument used when plotting histograms is bins. This argument is not in the pandas documentation so you do not see it when auto-tabbing for documentation. I have added bins=10 as an argument to the dataframe and series histogram function while also adding a line of documentation. 10 is the matplotlib default bin value so this is only making bins visible, not adding or changing any plotting functionality.

[jreback]

can you move the argument to the end of the arg list pls.

[TomAugspurger]

I'll look closer at this later, but the usual convention for our plotting functions is that any keyword arguments used in matplotlib's plot or hist or bar, etc. are passed in **kwargs.

Also you may want to use df.plot(kind='hist').

[zachcp]

@jreback updated argument list
@TomAugspurger I think this is import documentation for histograms. You can use df.plot(kind="hist") but I typically use series.hist(bins=20). For historical reasons or otherwise dataframes and series can call hist directly so having histogram-specific information there makes sense to me. I can certainly see why you wouldn't want bins in df.plot().

[TomAugspurger]

I misspoke earlier. hist isn't even a supported kind for df.plot()

[jreback]

@TomAugspurger what should we do with this?

[zachcp]

I'm happy whatever you decide to do. I just rebased against upstream/master in case its a go.

[TomAugspurger]

The downside is that this will be inconsistent with all our other plotting functions, where the arguments that go directly to the maptlotlib function are passed in the **kwargs section. But, it's good for discoverability, so let's merge.

@zachcp could you move the bins down in the docstring to match the order in the function signature?

Looks like you had a test fail too.

[zachcp]

made a couple minor fixes to padd test (need to pass bins as arg) and added two small tests.

[TomAugspurger]

OK thanks.

Don't need a release notes entry since this is really just a doc change. There isn't any change to the API

[TomAugspurger]

Oops, looks like we missed a call to hist in hist_frame. Not surprising given how spread out the code is. I've got a PR coming

[TomAugspurger]

merged #6935