如何解决matplotlib多子图全局总标题不显示、统一设置图例的问题

问题1：总标题不显示 #

原因是plt.suptitle()的调用位置错误，你将它放在了plt.subplots()之前，而plt.subplots()会创建新的画布，导致之前设置的总标题被覆盖。将plt.suptitle()移到plt.subplots()之后即可，同时可以调整y参数控制总标题的垂直位置，避免和子图标题重叠。

问题2：多子图共用统一图例 #

你需要先在每个子图的plot方法中添加legend=False参数关闭单张子图的图例，所有子图绘制完成后，提取任意一个子图的图例句柄和标签，在全局画布上添加统一图例即可，同时可以调整画布边距给图例留出空间。

import pandas as pd
import matplotlib.pyplot as plt

# 虚拟数据集生成
d = {
    'SeniorCitizen': [0,1,0,0,0,0,0,1,0,1,1,0,0,0,0,0,0,1,0,0,0,0,0,0,0] , 
    'CollegeDegree': [0,0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,1,1,1,1] , 
    'Married': [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1] , 
    'FulltimeJob': [1,1,1,1,1,0,0,0,1,1,1,1,1,1,1,1,1,0,0,1,1,0,0,0,1] , 
    'DistancefromBranch': [7,9,14,20,21,12,22,25,9,9,9,12,13,14,16,25,27,4,14,14,20,19,15,23,2] , 
    'ReversedPayment': [0,0,0,0,1,0,1,0,0,0,0,0,0,0,0,0,1,0,1,0,0,1,0,1,0] }
CarWash = pd.DataFrame(data = d)

categoricals = ['SeniorCitizen','CollegeDegree','Married','FulltimeJob','ReversedPayment']
numerical = ['DistancefromBranch']
CarWash[categoricals] = CarWash[categoricals].astype('category')

# 绘制频数子图
nrow = 1
ncol = len(categoricals[:-1])
figure1, axes1 = plt.subplots(nrow, ncol, figsize = (40,10))
# 总标题移到画布创建之后，调整y参数避免重叠
figure1.suptitle("Distribution of target variable across the categorical variables - frequencies #", fontsize=35, y=1.05)

for i,ax in zip(categoricals[:-1], axes1.flatten()):
    # 绘制频数，关闭单图legend
    CarWash.groupby([i,'ReversedPayment']).size().reset_index().pivot(index = i,columns = 'ReversedPayment').plot(kind = 'bar', stacked = True, ax = ax, sharey=True, legend=False)
    ax.tick_params(axis='both', labelsize = 25, labelrotation = 0)   
    ax.set_title(i, fontsize = 30)
    ax.set_xlabel("")
    # 标记数据
    for p in ax.patches:       
        x_adjust = 0.25
        value = p.get_height()
        X = p.get_x() + x_adjust
        Y = p.get_y() + p.get_height()/2
        XY = (X,Y)
        if value != 0:
            ax.annotate(int(value), XY, fontsize = 25)

# 提取图例句柄和标签，添加统一图例
handles, labels = axes1[0].get_legend_handles_labels()
# 调整标签名称为更易读的形式
labels = ['ReversedPayment=0', 'ReversedPayment=1']
figure1.legend(handles, labels, loc='upper right', bbox_to_anchor=(1.08, 0.95), fontsize=25)
# 调整画布边距，给图例留出空间
figure1.tight_layout()

# 绘制占比子图
figure2, axes2 = plt.subplots(nrow, ncol, figsize = (40,10))
figure2.suptitle("Distribution of target variable across the categorical variables - proportions %", fontsize=35, y=1.05)

for i,ax in zip(categoricals[:-1], axes2.flatten()):
    # 绘制占比，关闭单图legend
    CarWash.groupby([i,'ReversedPayment']).size().reset_index().pivot(index = i, columns = 'ReversedPayment').apply(lambda x: x/x.sum(),axis=1).plot(kind = 'bar', stacked = True, ax = ax, legend=False)
    ax.tick_params(axis='both', labelsize = 25, labelrotation = 0)   
    ax.set_title(i, fontsize = 30)
    ax.set_xlabel("")
    # 标记数据
    for p in ax.patches:       
        x_adjust = 0.25/3
        value = p.get_height()
        X = p.get_x() + x_adjust
        Y = p.get_y() + p.get_height()/2
        XY = (X,Y)
        if value != 0:
            ax.annotate(str(round(value*100,1)) + "%", XY, fontsize = 25)

# 占比图添加统一图例
figure2.legend(handles, labels, loc='upper right', bbox_to_anchor=(1.08, 0.95), fontsize=25)
figure2.tight_layout()

plt.show()

• 如果需要频数、占比两组子图共用同一个总标题，可以把两个子图放在同一个画布的上下两行，调整plt.subplots的nrow=2即可，逻辑和上述代码一致。
• 图例的位置可以通过调整bbox_to_anchor的参数自定义，也可以放在画布底部，避免占用横向展示空间。

内容的提问来源于stack exchange，提问作者Piyush