Matplotlib双面板分组柱状图偏移及X轴标签错乱问题排查

问题诊断 #

你的代码出现三个核心问题：

1. multiplier变量未重置：第一个面板绘制时multiplier累加至4，第二个面板继续使用这个值，导致柱子偏移量（offset = width * multiplier）过大，直接把柱子推到下一个类别区域，造成错位。
2. X轴标签无序：combos使用了Python集合（{}），集合是无序结构，Matplotlib读取时会打乱标签顺序，导致X轴标签混乱。
3. 刻度位置计算错误：每个类别下有4个分组柱子，原代码把刻度放在x + width位置，没有对准分组的中间，视觉上会出现错位。

修正后的代码 #

import matplotlib
import matplotlib.pyplot as plt
import numpy as np

### Create Plot ###
matplotlib.rc('xtick', labelsize=22) 
matplotlib.rc('ytick', labelsize=22)
    
# 改用列表保证标签顺序
combos = ['1 Product', '2 Product', '3 Product','4 Product', '5 Product', '6 Product','7 Product']
    
top30_rmse = {'RMSE Cold Season': [4.91726910421414,3.92792247654507,3.53717233977146,3.32461948965479,3.19029758875898,3.0975153495569,3.02950308430467],
             'RMSE Warm Season': [4.436715943,3.767955663,3.516888357,3.384377437,3.302319607,3.246462314,3.205968468]}

top30_corr = {'Correlation Cold Season': [0.639808613,0.71965916,0.75785035,0.780044305,0.794336664,0.80422278,0.811432104],
            'Correlation Warm Season': [0.888086922,0.936554339,0.952068621,0.958842973,0.962550154,0.964870985,0.966455676]}

depth_rmse = {'RMSE Cold Season': [3.564576325,3.000595232,2.787357328,2.6743701,2.60422583,2.556393944,2.521672889],
               'RMSE Warm Season': [4.706464351,4.416909958,4.316077216,4.264766936,4.233682287,4.212831767,4.197875134]}

depth_corr = {'Correlation Cold Season': [0.822695625,0.855445131,0.868612894,0.875592404,0.87985752,0.882709683,0.884740798],
            'Correlation Warm Season': [0.858527248,0.880016071,0.887233404,0.890810139,0.892940312,0.894352646,0.89535723]}

x = np.arange(len(combos))  # 基于列表长度生成x，适配元素数量变化
width = 0.2  # 调整宽度，4个柱子总宽度0.8，避免相邻类别重叠
fig, (ax1, ax3) = plt.subplots(nrows=2,ncols=1,sharex='all',figsize=(20,20))

## Top-30cm ##
multiplier = 0  # 每个面板开始前重置multiplier
# RMSE 绘制
for stat, group in top30_rmse.items():
    offset = width * multiplier
    rects = ax1.bar(x + offset, group, width, label=stat)
    multiplier += 1
# 相关性绘制
ax2 = ax1.twinx()
for stat, group in top30_corr.items():
    offset = width * multiplier
    rects = ax2.bar(x + offset, group, width, hatch='x', label=stat)
    multiplier += 1    
# 设置X轴刻度到分组中心
ax1.set_xticks(x + width*1.5, combos)
ax1.set_ylim(0,10)
ax2.set_ylim(0.5,1)
# 添加图例提升可读性
ax1.legend(loc='upper left')
ax2.legend(loc='upper right')

## Depth ##
multiplier = 0  # 重置multiplier
# RMSE 绘制
for stat, group in depth_rmse.items():
    offset = width * multiplier
    rects = ax3.bar(x + offset, group, width, label=stat)
    multiplier += 1
# 相关性绘制
ax4 = ax3.twinx()
for stat, group in depth_corr.items():
    offset = width * multiplier
    rects = ax4.bar(x + offset, group, width, hatch='x', label=stat)
    multiplier += 1                        
# 设置X轴刻度到分组中心
ax3.set_xticks(x + width*1.5, combos)
ax3.set_ylim(0,10)
ax4.set_ylim(0.5,1)
# 添加图例
ax3.legend(loc='upper left')
ax4.legend(loc='upper right')

plt.tight_layout()
plt.show()

关键修改点 #

• 将combos从集合改为列表，确保X轴标签按1-7 Product的顺序显示。
• 每个面板绘制前重置multiplier为0，保证第二个面板的柱子从正确的偏移量开始绘制。
• 调整柱子宽度为0.2，4个柱子总宽度0.8，避免相邻类别重叠；同时将X轴刻度设置为x + width*1.5，对准每个分组的中心位置。
• 为每个轴添加图例，提升图表可读性。

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