使用brokenaxis()时x轴刻度标签偏移且缺失的问题求助

I totally get your frustration—when using brokenaxes, the x-tick labels don’t behave like they do with regular Matplotlib axes, which is why you’re only seeing 5 labels with the first one missing. Here’s how to fix this and get all 9 labels to display exactly like your unbroken plot:

The Root Cause #

brokenaxes creates multiple sub-axes under the hood instead of a single axis. By default, only the bottom sub-axis shows x-ticks, and it auto-adjusts the tick count, leading to missing or truncated labels. To fix this, you need to explicitly configure each sub-axis to use your full set of ticks and labels.

Solution Code #

Here’s your modified plotting code that retains all x-tick labels matching your original unbroken plot:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from brokenaxes import brokenaxes

# Computing dataframe (unchanged from your code)
column1 = np.array(('Grünland','KA','MA','SG','SW','WG','WR','WW','ZR'))
column2 = np.array((1332.9,-429.7,-6041.6,-686.5,-72.4,-37.5,-4224.6,-4208.8,-3955.8))
column3 = np.array((1332.9,42.7,-1018.7,-32.7,21.7,-327.2,-3117.4,-15431.5,206.2))
data1 = np.column_stack((column1,column2,column3))
data = pd.DataFrame(data1,columns=['Frucht','SUM19','SUM20'])
data['SUM19'] = pd.to_numeric(data['SUM19'])
data['SUM20'] = pd.to_numeric(data['SUM20'])
labels = data['Frucht']
x = np.arange(len(labels))  # the label locations
width = 0.35  # the width of the bars

# Plot with broken y-axis (modified part)
fig = plt.figure()
bax = brokenaxes(ylims=((-16000,-15000), (-6500, 1500)), hspace=.05)
rects1 = bax.bar(x - width/2, data['SUM19'], width, label='2019')
rects2 = bax.bar(x + width/2, data['SUM20'], width, label='2020')

# Configure x-ticks for all sub-axes
for ax in bax.axes:
    # Set exact tick positions and your custom labels
    ax.set_xticks(x)
    ax.set_xticklabels(labels)
    # Expand x-axis limits to prevent label cutoff
    ax.set_xlim(-0.5, len(x) - 0.5)
    # Hide labels on upper sub-axes to avoid duplicate text
    if ax != bax.axes[-1]:
        ax.tick_params(labelbottom=False)

plt.legend()
plt.show()

Key Changes Explained #

1. Iterate over all sub-axes: bax.axes gives you access to every axis created by brokenaxes. Setting ticks and labels for each ensures alignment across the broken plot.
2. Set x-axis limits: ax.set_xlim(-0.5, len(x)-0.5) makes sure the first and last labels aren’t cut off by the plot edges.
3. Hide duplicate labels: We only keep labels on the bottom sub-axis (bax.axes[-1]) to avoid overlapping text on upper axes while keeping ticks aligned.

This should make your broken-axis plot’s x-tick labels match exactly with your original unbroken plot!

内容的提问来源于stack exchange，提问作者user13897961