将数据帧合并到模型矩阵:r

import pandas as pd
from sklearn.preprocessing import OneHotEncoder

# 示例数据帧
df = pd.DataFrame({
    'feature1': [1, 2, 3],
    'category': ['A', 'B', 'A']
})

# 对分类变量进行独热编码
encoder = OneHotEncoder()
encoded_category = encoder.fit_transform(df[['category']]).toarray()

# 合并数值列和编码后的分类列
model_matrix = pd.concat([df[['feature1']], pd.DataFrame(encoded_category, columns=encoder.get_feature_names_out(['category']))], axis=1)

# 示例数据帧
df = pd.DataFrame({
    'feature1': [1, None, 3],
    'feature2': [4, 5, 6]
})

# 填充缺失值（例如用均值填充）
df.fillna(df.mean(), inplace=True)

# 转换为模型矩阵
model_matrix = df.values

# 示例数据帧
df1 = pd.DataFrame({'A': [1, 2], 'B': [3, 4]})
df2 = pd.DataFrame({'B': [5, 6], 'A': [7, 8]})

# 确保列顺序一致
df1 = df1[['A', 'B']]
df2 = df2[['A', 'B']]

# 合并数据帧
merged_df = pd.concat([df1, df2], axis=0)
model_matrix = merged_df.values