Plotting a pandas dataframe? about bokeh

[pzwang]

There isn't, but there's also no reason why we couldn't implement this pretty easily. I need to look into how df.plot works more closely, to see what we need to do to replicate its behavior. From what I can tell by looking at the documentation (http://pandas.pydata.org/pandas-docs/stable/visualization.html), by default it overlays the named columns in a single plot, and it can optionally facet into subplots (but that seems to require directly using Matplotlib to set up the subplot frames?).

Plot types are: line, scatter, bar/stacked bar, parallel coordinates, and boxplot. Does this pretty much cover it?

[jvns]

That's all the functionality I've used myself -- just tried the subplot option right this second. I think it would be great if it could just overlay the named columns in a single plot with a legend, to start.

[hughesadam87]

Hi guys,

This probably isn't helpful much, but I wanted to confirm that in my experience, there is no way in pandas to facet into subplots without setting them up beforehand in matplotlib with a call to plt.subplots(). Some advanced pandas plots like the scatter matrix will do this for you when they are called.

I wanted to point out that df.plot() returns an AxesSubplot object (eg <matplotlib.axes.AxesSubplot object at 0x38e31d0>), so maybe you'd just need to facet to this? I don't understand what would be involved of course...

[damianavila]

We have duplicated issues about pandas support.
I will close this one and link it to a more recent discussion living in the issue #140.
I will also transfer the milestone and the labels to the other issue.