How to get rid of superfluous x-axis date tick marks with pandas plot?-CodePudding

Using pandas' plot method, I am trying to plot the columns in a pandas DataFrame whose index consists of dates with regular spacing except the last one.

For some reason, superfluous tick marks appear on the x-axes of the plots. How can I get rid of these superfluous tick marks while continuing to use pandas' plot method?

I am pasting below a self-contained code that reproduces the problem and a screenshot of the output in which I have marked some of the superfluous tick marks. I am using Python 3.10.4, pandas 1.4.2, Numpy 1.22.3, and Matplotlib 3.5.2.

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import datetime as dt
 
x = np.arange(-3, 3, 0.1)
dates = pd.date_range(end=dt.date(2022, 9, 30), freq="Q", periods = len(x), name="date")
 
y1 = np.sin(x)
y2 = np.cos(x)

trig = pd.DataFrame({"sin": y1, "cos": y2}, index = dates)

extra_date = pd.date_range(end=dt.date(2022, 10, 19), freq="D", periods = 1, name="date")
extra = pd.DataFrame({"sin": np.sin(2.91), "cos": np.cos(2.91)}, index = extra_date)
trig = pd.concat([trig, extra])
 
fig, axes = plt.subplots(2, 1, figsize=(15, 7), sharex=True) 
trig.sin.plot(ax=axes[0], grid=True, style=["o-"], rot=0, title="Sin")
trig.cos.plot(ax=axes[1], grid=True, style=["o-"], rot=0, title="Cos")

for ax in axes:
    for label in ax.xaxis.get_ticklabels():
        label.set_ha('center')

plt.show()

The problem disappears when I allow each plot to have a separate x-axis by removing the sharex=True from plt.supbplots() but I don't want the plots to have separate x-axes and I wish to continue using pandas' plot method.

I have looked at various shared x-axis tick mark-related Q&As involving dates, Matplotlib, and pandas on Stack Overflow but could not find anyting solving this problem without requiring me to give up on pandas' plot method.

CodePudding user response：

Modify the for loop as

for ax in axes:
    ax.minorticks_off()
    for label in ax.xaxis.get_ticklabels():
        label.set_ha('center')