Duplicate of #1918
Basically the problem is if you want to do a bar plot with time series
df.plot(kind='bar')
it will create a label for every date entry instead one clean label and more dates when you zoom in.
Comment From: TomAugspurger
Usually bar charts are used for non-continuous data. With non-continuous data there may not be any way to infer the label for a suppressed tick. So I think that displaying every tick is a good default for kind='bar'
That said, it might make sense treat a bar chart with a DatetimeIndex differently, since in that case there is an ordering to the data.
Comment From: dfrusdn
This would be used if you are plotting volume for a stock in a time series. It should be easy to implement too.
Comment From: ghost
I read the description and it's still unclear to me what the problem is and what the desired behavior is. If you could elaborate that would help. So would screenshots.
This sounds like something we might at least accept a PR for.
Comment From: jreback
@dfrusdn if you are interested in contributing to the, pls submit a PR
Comment From: jorisvandenbossche
Closing as a duplicate of #1918