Also, I notice that in the docstrings there are Valid kwargs are and Valid *kwargs* are, should we change make kwargs italic in these cases to keep consistency? or the other way around?

@timhoffm Thanks for the reviewing! I added two more commits:

• Commit 27da050:
  Change those:

  :class:`~matplotlib.lines.Line2D`, :class:`matplotlib.lines.Line2D`, `~.Line2D`
  

  to `.Line2D`

• Commit 87ff17c: kwargs are, Valid kwargs are -> Valid keyword arguments are:

Sorry @bingyao this now needs a rebase.... Please squash your commits as well.

@jklymak Hi, I just finished rebasing and squashing.

I resolved it. The online resolver is actually pretty good.

BTW, it was "string"->"str" that caused the conflicts here.

Sorry for the mis-commit there...

@timhoffm @jklymak It's fine. I totally understand the situation. 😄Just did another rebasing and squashing.

I think the testing failure of lib/matplotlib/tests/test_axes.py::test_bar_pandas is due to the setting in:
lib/matplotlib/testing/conftest.py:

try:
 from pandas.plotting import (
 deregister_matplotlib_converters as deregister)
 deregister()
except ImportError:
 pass

Although I am not very clear why deregister_matplotlib_converters() is needed in here as an overall setting.

We deregister the pandas converters because we want pandas to test their own converter and we will test ours. It seems the new version of pandas has broken out tests so suggest we pin away from it until we sort the problem out

@jklymak Thanks! And yes, it's the pandas 0.25.0 that causes the failure, 0.24.2 was fine.

Can we change the test_bar_pandas(pd) to following to avoid this failure?

def test_bar_pandas(pd):
 # Smoke test for pandas
 pd.plotting.register_matplotlib_converters()
 fig, ax = plt.subplots()
 df = pd.DataFrame(
 {'year': [2018, 2018, 2018],
 'month': [1, 1, 1],
 'day': [1, 2, 3],
 'value': [1, 2, 3]})
 df['date'] = pd.to_datetime(df[['year', 'month', 'day']])
 monthly = df[['date', 'value']].groupby(['date']).sum()
 dates = monthly.index
 forecast = monthly['value']
 baseline = monthly['value']
 ax.bar(dates, forecast, width=10, align='center')
 ax.plot(dates, baseline, color='orange', lw=4)
 pd.plotting.deregister_matplotlib_converters()

i.e. register at beginning and deregister at the end.

Sorry for the delay. And thanks @timhoffm @QuLogic for the references and reminding!

@jklymak Hi, I think the Needs rebase label can be removed at this time, right? 😄