如何处理熊猫DataFrame中的缺失值？

import pandas as pd

# 创建一个示例 DataFrame
data = {
    'A': [1, 2, None, 4],
    'B': [5, None, 7, 8],
    'C': [None, 10, 11, 12]
}
df = pd.DataFrame(data)

# 检查缺失值
missing_values = df.isnull()
print(missing_values)

# 删除包含缺失值的行
df_cleaned = df.dropna()

# 删除包含缺失值的列
df_cleaned = df.dropna(axis=1)

# 使用均值填充缺失值
df_filled = df.fillna(df.mean())

# 使用特定值填充缺失值
df_filled = df.fillna(0)

# 使用前一个有效值填充缺失值
df_filled = df.fillna(method='ffill')

# 使用后一个有效值填充缺失值
df_filled = df.fillna(method='bfill')

# 线性插值
df_interpolated = df.interpolate()

# 多项式插值
df_interpolated = df.interpolate(method='polynomial', order=2)

from sklearn.ensemble import RandomForestRegressor

# 假设我们要填充列 'A' 中的缺失值
missing_indices = df[df['A'].isnull()].index

# 移除缺失值行
df_no_missing = df.dropna(subset=['A'])

# 训练模型
X_train = df_no_missing.drop(columns=['A'])
y_train = df_no_missing['A']
model = RandomForestRegressor()
model.fit(X_train, y_train)

# 预测缺失值
X_missing = df.loc[missing_indices].drop(columns=['A'])
predicted_values = model.predict(X_missing)

# 填充缺失值
df.loc[missing_indices, 'A'] = predicted_values