matplotlib date x-ticks bug bar vs area graph

Asked Oct 27 '22 at 14:58

Active Oct 27 '22 at 15:04

Viewed 27 times

I run the below code for kind = "bar" and kind = "area" and get incorrect dates for "bar". Could this bug be confirmed and perhaps a suggestion to mitigate it?

import datetime as dt
import pandas as pd
import matplotlib.dates as mdates

kind = "bar" ### "area" works
data = {
    dt.date(2022, 10, 4): {"x": 123, "y": 234, "z": 345},
    dt.date(2022, 10, 5): {"x": 678, "y": 789, "z": 980},
    dt.date(2022, 10, 6): {"x": 987, "y": 876, "z": 765},
}
plt = pd.DataFrame(data).T.plot(
    kind = kind,
    rot = "vertical",
)
ax = plt.axes
ax.xaxis.set_major_formatter(mdates.DateFormatter("%Y-%m-%d"))
ax.legend(bbox_to_anchor = (1, 1))

edited Oct 27 '22 at 15:04

asked Oct 27 '22 at 14:58

Shuri2060

Note "%Y-%m-%d" as choice of date format was just to showcase the bug, I want to have a custom format – Shuri2060 Oct 27 '22 at 14:59
Not a bug. The xticks are `0`, `1`, `2` for your bar chart. – BigBen Oct 27 '22 at 15:01
Could you explain further? If I comment out `x.xaxis.set_major_formatter(mdates.DateFormatter("%Y-%m-%d"))`, I get the correct x-ticks again for "bar" – Shuri2060 Oct 27 '22 at 15:03
Those are the xticklabels, not the xticks, as can be seen with `ax.get_xticklabels()` vs `ax.get_xticks()`. – BigBen Oct 27 '22 at 15:03
So what exactly should I be passing that DateFormatter to? – Shuri2060 Oct 27 '22 at 15:11
2

https://stackoverflow.com/questions/30133280/pandas-bar-plot-changes-date-format – BigBen Oct 27 '22 at 15:14

matplotlib date x-ticks bug bar vs area graph

0 Answers0