我不知道如何进行预测，因为我的训练数据和测试数据不同，我不知道如何处理这些差异和缺失值。以下是我的代码：

import pandas as pdfrom sklearn.model_selection import train_test_split# 读取数据X = pd.read_csv('../input/train.csv', index_col='Id') X_test = pd.read_csv('../input/test.csv', index_col='Id')# 删除目标值缺失的行，并将目标值与预测变量分离X.dropna(axis=0, subset=['SalePrice'], inplace=True)y = X.SalePriceX.drop(['SalePrice'], axis=1, inplace=True)# 为了简化，我们将删除包含缺失值的列cols_with_missing = [col for col in X.columns if X[col].isnull().any()] X.drop(cols_with_missing, axis=1, inplace=True)X_test.drop(cols_with_missing, axis=1, inplace=True)# 从训练数据中分离出验证集X_train, X_valid, y_train, y_valid = train_test_split(X, y,                                                      train_size=0.8, test_size=0.2,                                                      random_state=0)### 独热编码# 将进行独热编码的列low_cardinality_cols = [col for col in object_cols if X_train[col].nunique() < 10]# 将从数据集中删除的列high_cardinality_cols = list(set(object_cols)-set(low_cardinality_cols))from sklearn.preprocessing import OneHotEncoder# 复制数据以避免在填补时改变原始数据X_train_new = X_train.copy()X_valid_new = X_valid.copy()# 对低基数列应用独热编码器OH_encoder = OneHotEncoder(handle_unknown='ignore', sparse=False)OH_cols_train = pd.DataFrame(OH_encoder.fit_transform(X_train_new[low_cardinality_cols]))OH_cols_valid = pd.DataFrame(OH_encoder.transform(X_valid_new[low_cardinality_cols]))# 恢复在独热编码过程中丢失的索引OH_cols_train.index = X_train_new.indexOH_cols_valid.index = X_valid_new.index# 删除我们将用独热编码替换的分类列（对象调用因为# 我们也想删除高基数列）num_X_train = X_train_new.drop(object_cols, axis=1)num_X_valid = X_valid_new.drop(object_cols, axis=1)# 将独热编码的列添加到数值特征中OH_X_train = pd.concat([num_X_train, OH_cols_train], axis=1) OH_X_valid = pd.concat([num_X_valid, OH_cols_valid], axis=1) 

到目前为止一切正常。但是当我想进行预测时，失败了。以下是我的代码：

# 复制数据以避免在填补时改变原始数据X_train_new = X_train.copy()X_valid_new = X_valid.copy()# 🍕X_test_new = X_test.copy()# 为了简化，我们将删除包含缺失值的列cols_with_missing = [col for col in X_test_new.columns if X_test_new[col].isnull().any()] X_test_new.drop(cols_with_missing, axis=1, inplace=True)# 测试数据中的分类列new_object_cols = [col for col in X_test_new.columns if X_test_new[col].dtype == "object"]# 将进行独热编码的列new_low_cardinality_cols = [col for col in new_object_cols if X_test_new[col].nunique() < 10]# 将从数据集中删除的列new_high_cardinality_cols = list(set(new_object_cols)-set(new_low_cardinality_cols))OH_cols_test = pd.DataFrame(OH_encoder.transform(X_test_new[new_low_cardinality_cols]))

这是我的错误信息：

ValueError: The number of features in X is different to the number of features of the fitted data. The fitted data had 24 features and the X has 19 features.

回答：

您在训练/测试数据中出现形状不匹配的一个原因是您在进行训练/测试分割之后创建了新的分类变量。最可能的情况是，您有一些类别只出现在训练或测试分割中，因此形状最终不匹配。

我建议将以下训练/测试分割行：

X_train, X_valid, y_train, y_valid = train_test_split(X, y, train_size=0.8, test_size=0.2, random_state=0)

移动到预处理步骤的末尾，以避免这个问题，并且确保以完全相同的方式预处理训练/测试数据。