如何解决'Axes' object has no attribute 'is_first_col'报错？

完全按照练习册编写的代码如下：

from sklearn.model_selection import train_test_split, cross_val_score, StratifiedKFold
from matplotlib import pyplot
from sklearn.preprocessing import StandardScaler
import pandas as pd
import matplotlib.pyplot as plt


df = pd.read_csv('D:\pima-indians-diabetes3.csv')

X = df.iloc[:, 0:8]
y = df.iloc[:, 8]

ss = StandardScaler()
scaled_X = pd.DataFrame(ss.fit_transform(X), columns = X.columns)
fig, ax = plt.subplots(1, 2, figsize = (12, 4))
X.plot(kind='kde', title = 'Raw data', ax=ax[0])
scaled_X.plot(kind='kde', title = 'StandardScaler', ax=ax[1])
plt.show()

运行时在以下两行代码处报错：

X.plot(kind='kde', title = 'Raw data', ax=ax[0])
scaled_X.plot(kind='kde', title = 'StandardScaler', ax=ax[1])

报错信息：Axes' object has no attribute 'is_first_col'

这个错误是pandas与matplotlib版本不兼容导致的：is_first_col是pandas 1.5+版本中用于绘图布局的内部属性，若你的matplotlib版本过低（低于3.6），其Axes对象不包含该属性，就会触发此报错。

方案一：升级matplotlib到兼容版本 #

执行以下命令升级matplotlib，确保版本≥3.6：

pip install --upgrade matplotlib

方案二：调整代码适配旧版matplotlib #

如果暂时无法升级依赖，可以修改绘图逻辑，绕开pandas对is_first_col的依赖：

from sklearn.model_selection import train_test_split, cross_val_score, StratifiedKFold
from matplotlib import pyplot
from sklearn.preprocessing import StandardScaler
import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('D:\pima-indians-diabetes3.csv')

X = df.iloc[:, 0:8]
y = df.iloc[:, 8]

ss = StandardScaler()
scaled_X = pd.DataFrame(ss.fit_transform(X), columns = X.columns)

# 手动创建子图并分别绘制
plt.figure(figsize=(12,4))

plt.subplot(1,2,1)
X.plot(kind='kde', title='Raw data', ax=plt.gca())

plt.subplot(1,2,2)
scaled_X.plot(kind='kde', title='StandardScaler', ax=plt.gca())

plt.tight_layout()
plt.show()

内容的提问来源于stack exchange，提问作者Bella Lee