DF Plots: Error bars don't allow to set styles

[Zaharid]

With this code the style spec 'o' is ignored when plotting error bars (#5638):

In [25]: from numpy.random import randn

In [26]: import pandas as pd

In [27]: df = pd.DataFrame(randn(8,3), columns = "A B yerr".split())

In [28]: df.plot(x='A', y='B',style='o' ,yerr='yerr')
Out[28]: <matplotlib.axes.AxesSubplot at 0x5259cd0>

In [29]: pd.__version__
Out[29]: '0.13.1-791-g07f6d46'

[TomAugspurger]

Thanks for the report. Any fix for this will have to follow #6834 which is refactoring the error bar code (doesn't fix this though). The style kwarg is checked for when plotting the error bars and removed since it isn't supported, but I think that it should just not style the actual error bars. The plot itself should still be styled.

[cwwalter]

I found this bug while trying to solve this problem. I'm new to trying to plot with Pandas. Although the style command does not work, you can still do what you want by using the fmt command like so:

data.query('config==0').plot('numElectrons', 'stdX', yerr='errX', fmt='ko', label='Perfect X')

[dashesy]

fmt workaround does not work if subplots is used

[MarineF]

fmt workaround also does not work when we need to modify line property such as in:
df.plot(style=['k:', 'k--', 'k-'], yerr=error_df)

[Octoshape]

Is this ever going to be fixed?