使用 sk-learn 根据字符串特征预测数值特征

Question

我正在尝试预测“ Full_Time_Home_Goals”列（功能）。我遵循了Kaggle 的例子。该代码适用于不同的维度，如我的示例（测试数据中的 419 行和训练数据中的 892 行）

我收到的错误是：

1. 代码如下：

   回溯（最近一次调用最后一次）：文件“C:/Users/harsh/PycharmProjects/Kaggle-Machine Learning from Start to Finish with Scikit-Learn/EPL Predicting.py”，第 98 行，预测 = clf.predict(data_test. drop('Id', axis=1)) 文件“C:\Users\harsh\PycharmProjects\GitHub\venv\lib\site-packages\sklearn\ensemble_forest.py”，第 629 行，预测 ValueError：特征数量模型必须与输入匹配。模型 n_features 为 4，输入 n_features 为 2

2. 将代码更改为 后predictions = clf.predict(data_test.drop('Id', axis=1)) to predictions = clf.predict(X_test)，错误为：

    raise ValueError(msg) ValueError: array length 37921 does not match index length 380

我该如何解决这个问题？

我使用的数据集可以在这里找到

Answer 1

以下是您经过测试且完全有效的代码：data_train = pd.read_csv(r"train.csv")data_test = pd.read_csv(r"test.csv")columns = ['Id', 'HomeTeam', 'AwayTeam', 'Full_Time_Home_Goals']col = ['Id', 'HomeTeam', 'AwayTeam']data_test = data_test[col]data_train = data_train[columns]data_train = data_train.dropna()data_test = data_test.dropna()data_train['Full_Time_Home_Goals'] = data_train['Full_Time_Home_Goals'].astype(int)from sklearn import preprocessingdef encode_features(df_train, df_test):    features = ['HomeTeam', 'AwayTeam']    df_combined = pd.concat([df_train[features], df_test[features]])    for feature in features:        le = preprocessing.LabelEncoder()        le = le.fit(df_combined[feature])        df_train[feature] = le.transform(df_train[feature])        df_test[feature] = le.transform(df_test[feature])    return df_train, df_testdata_train, data_test = encode_features(data_train, data_test)print(data_train.head())print(data_test.head())# X_all would contain all columns required for prediction and y_all would have that one columns we want to predicty_all = data_train['Full_Time_Home_Goals']X_all = data_train.drop(['Full_Time_Home_Goals'], axis=1)from sklearn.model_selection import train_test_splitnum_test = 0.20  # 80-20 splitX_train, X_test, y_train, y_test = train_test_split(X_all, y_all, test_size=num_test, random_state=23)from sklearn.ensemble import RandomForestClassifierfrom sklearn.metrics import make_scorer, accuracy_scorefrom sklearn.model_selection import GridSearchCV# Using Random Forest and using parameters that we definedclf = RandomForestClassifier()parameters = {'n_estimators': [4, 6, 9],              'max_features': ['log2', 'sqrt', 'auto'],              'criterion': ['entropy', 'gini'],              'max_depth': [2, 3, 5, 10],              'min_samples_split': [2, 3, 5],              'min_samples_leaf': [1, 5, 8]              }acc_scorer = make_scorer(accuracy_score)grid_obj = GridSearchCV(clf, parameters, scoring=acc_scorer)grid_obj = grid_obj.fit(X_train, y_train)clf = grid_obj.best_estimator_clf.fit(X_train, y_train)predictions = clf.predict(X_test)print(accuracy_score(y_test, predictions))ids = data_test['Id']predictions = clf.predict(data_test)df_preds = pd.DataFrame({"id":ids, "predictions":predictions})df_preds   Id  HomeTeam  AwayTeam  Full_Time_Home_Goals0   1        55       440                     31   2       158       493                     22   3       178       745                     13   4       185       410                     14   5       249        57                     2       Id  HomeTeam  AwayTeam0  190748       284        541  190749       124       4412  190750       446        573  190751       185       6374  190752       749       4820.33213786556261704id  predictions0   190748  11   190749  12   190750  13   190751  14   190752  1... ... ...375 191123  1376 191124  1377 191125  1378 191126  1379 191127  1380 rows × 2 columns