Matplotlib tick-label warning with pandas 3.1.0 #582

I got this warning after upgrading to pandas 3.1.0:

/path/to/python314/lib64/python3.14/site-packages/quantstats/_plotting/core.py:266: UserWarning: set_ticklabels() should only be used with a fixed number of ticks, i.e. after set_ticks() or using a FixedLocator. Otherwise, ticks may be mislabeled.
  ax.set_xticklabels(df.index.year)

Environment: Python 3.14, pandas 3.1.0, matplotlib 3.11.1, quantstats 0.0.86

Example

import pandas as pd
import quantstats as qs

r = pd.Series([0.1, -0.05], index=pd.date_range("2025", periods=2, freq="YE"))
qs.plots.yearly_returns(r, show=False)

I asked ChatGPT to suggest a fix and it said the following:

In quantstats/_plotting/core.py, replace:

Old:

df.plot(kind="bar", ax=ax, color=colors)

New:

df.set_axis(df.index.astype(str)).plot(kind="bar", ax=ax, color=colors)

No other changes required.

Pandas 3.1 introduced dynamic date ticks for datetime-indexed bar plots. Converting the plotting index to strings restores categorical tick positioning while preserving the original DataFrame and existing year-labeling logic.

This eliminates the warning without changing the plotted returns or chart appearance.