如何仅用Pandas为条形图添加重复字符串占比标签

我有如下的Pandas DataFrame（可复现代码如下），已经生成了条形图，现在想**仅使用Pandas（允许预先硬编码）**为条形图添加重复字符串的百分比标签。目前已有Matplotlib的实现方案，但优先考虑纯Pandas实现，若无法实现再使用Matplotlib。

可复现代码：

import pandas as pd
import io   

TESTDATA="""All_services
All_services
Rehosting applications to AWS
Replacing flexible functionalities
Unaltered replatforming of underlying code structure, functionalities, features
Rebuilding broken applications/software segments
Optimize existing use of cloud(Cost saving)
Expand use of containers
Move on prem servers to Sass
Expanding public clouds
Implemenation CI/CD to clouds
Migration Evaluator
AWS Migration Hub 
AWS Application Discovery Services
AWS Landing Zone 
AWS Control Tower
AWS Management and Governance
AWS Database Migration Services
AWS Server Migration Service
AWS Database Migration Service
AWS Application Discovery Service
AWS Direct Connect
DB Migrations
open-source databases to AWS. 
Oracle to Oracle
Oracle or Microsoft SQL Server to Amazon Aurora. 
Migrating fileservers to Amazon S3
migrating commercial RDBMS or MySQL.
Optimize existing use of cloud(Cost saving)
Expand use of containers
Move on prem servers to Sass
Expanding public clouds
Implemenation CI/CD to clouds
Migration Evaluator
AWS Migration Hub 
AWS Application Discovery Services
AWS Landing Zone 
DB Migrations
Cloud Migration Planning                                              
Replatforming Applications for Cloud                                      
Cloud Application Development Services                                
From Monolith to Microservices                                                 
Cloud Infrastructure Automation                                                
Implemenation CI/CD to clouds
DB Migrations
Optimize existing use of cloud(Cost saving)
Implemenation CI/CD to clouds
Migration Evaluator
AWS Migration Hub 
AWS Application Discovery Services
AWS Direct Connect
DB Migrations
open-source databases to AWS. 
Oracle to Oracle
Oracle or Microsoft SQL Server to Amazon Aurora. 
Migrating fileservers to Amazon S3
Optimize existing use of cloud(Cost saving)
Amazon S3 Transfer Acceleration 
AWS Snowball
AWS Direct Connect
EC2
AWS Server Migration Service
AWS Database Migration Service 
VMWare Cloud on AWS 
Optimize existing use of cloud(Cost saving)
Cloud Application Development Services                                
From Monolith to Microservices                                                 
Cloud Infrastructure Automation                                                
Implemenation CI/CD to clouds
DB Migrations
Optimize existing use of cloud(Cost saving)
Implemenation CI/CD to clouds
Migration Evaluator
Optimize existing use of cloud(Cost saving)
AWS Application Discovery Services
AWS Direct Connect
DB Migrations
Rebuilding broken applications/software segments
Optimize existing use of cloud(Cost saving)
Expand use of containers
Move on prem servers to Sass
Expanding public clouds
Implemenation CI/CD to clouds
AWS Management and Governance
AWS Database Migration Services
AWS Server Migration Service
AWS Database Migration Service
AWS Application Discovery Service
AWS Direct Connect
DB Migrations
"""

df = pd.read_csv(io.StringIO(TESTDATA), sep=";")
df = df.replace(r"^ +| +$", r"", regex=True)
df.All_services.value_counts().sort_values().plot(kind = 'barh',figsize=(25, 15),linewidth=4)

核心说明 #

纯Pandas无法直接给条形图添加自定义标签——Pandas的plot()方法只是封装了Matplotlib的绘图逻辑，本身没有提供添加标签的API。所以必须结合Matplotlib完成标签绘制，但我们可以用Pandas先完成所有数据计算，尽量减少非Pandas代码。

步骤1：用Pandas计算频次和百分比 #

先统计每个服务的出现次数，再算出对应的百分比：

# 统计频次并排序
counts = df['All_services'].value_counts().sort_values()
# 计算总数
total = counts.sum()
# 计算百分比（保留两位小数）
percentages = (counts / total * 100).round(2)
# 组合成要显示的标签文本（次数+百分比）
label_texts = [f"{count} ({pct}%)" for count, pct in zip(counts, percentages)]

步骤2：绘图并添加标签 #

用Pandas绘制条形图后，获取Matplotlib的坐标轴对象，逐个添加标签：

import matplotlib.pyplot as plt

# 绘制条形图，获取坐标轴实例
ax = counts.plot(kind='barh', figsize=(25, 15), linewidth=4)

# 遍历每个条形，添加标签
for idx, (text, count) in enumerate(zip(label_texts, counts)):
    # 调整文本位置，避免和条形重叠
    ax.text(count + 0.1, idx, text, va='center', fontsize=12)

# 自动调整布局，防止标签被截断
plt.tight_layout()
plt.show()

简化版：仅显示百分比 #

如果只需要展示百分比，也可以直接用百分比数据绘图并标注：

ax = percentages.plot(kind='barh', figsize=(25, 15), linewidth=4)
for idx, pct in enumerate(percentages):
    ax.text(pct + 0.1, idx, f"{pct}%", va='center', fontsize=12)
plt.tight_layout()
plt.show()

内容的提问来源于stack exchange，提问作者Bhargav