Question

def get_scores(model, model_name):


    model.fit(X_train, y_train)


    train_preds = model.predict(X_train)
    test_preds = model.predict(X_test)

    train_probs = model.predict_proba(X_train)
    test_probs = model.predict_proba(X_test)


    print('{} has training accuracy of: {}'.format(model_name, accuracy_score(y_train, train_preds)))     
    print('{} has test accuracy of: {}\n'.format(model_name, accuracy_score(y_test, test_preds)))
    print('{} has training log loss of: {}'.format(model_name, log_loss(y_train, train_probs)))
    print('{} has test log loss of: {}\n'.format(model_name, log_loss(y_test, test_probs)))

我想稍后再调用多个模型，我在这里调用其中之一：

lr_initial = LogisticRegression(n_jobs=-1)

initial_lr_train_preds, initial_lr_test_preds, initial_lr_train_probs, initial_lr_test_probs = get_scores(lr_initial, ' Preliminary Logistic Regression model')

因此，我想将每个模型的accuracy_score(y_train, train_preds)的所有值放入变量中。该怎么办

Answer 1

有几种编码方法可以实现您所需要的。例如，您可以使用{y_train：（train_preds，value）}形式的字典，其中y_train是键，而元组（train_preds，value）则是字典值

如何在一个函数的多个变量中分配多个值？

1 个答案: