Question

确切的错误是在行上发生的“ ValueError：标签数= 21与样本数= 82不匹配”

rf.fit(train_X, train_y)

代码如下：

df = pd.read_csv('C:\\git\\MetalRater\\Metal Sheet 2 - Sheet1 - TEST.csv', encoding="ISO-8859-1")

# The x and y are defined (x = features, y = y)
features = ["Emotion", "Solid", "Variety", "Length (mins)"]
y = df["RL"]

train_X, test_X, train_y, test_y = train_test_split(df[features], y, test_size=0.2, random_state=0)

print(len(train_X))
print(len(train_y))

def find_n_estimators(train_X, train_y, test_X, test_y):
    accuracy_forest_base = 0
    for i in range(10, 1000, 10):
        rf = RandomForestRegressor(random_state = 0, n_estimators = i)
        rf.fit(train_X, train_y)
        predictions_forest = rf.predict(test_X)
        for i in range(len(predictions_forest)):
            predictions_forest[i] = round(predictions_forest[i],0)
        accuracy_forest = accuracy_score(test_y, predictions_forest)
        if accuracy_forest > accuracy_forest_base:
            accuracy_forest_base = accuracy_forest
            n_est = i
        else:
            break
    return n_est

打印语句确认两者的长度均为82。

编辑：按照下面的要求，我打印了以下内容：

print(np.shape(train_X)[0])
print(np.shape(train_y)[0])

结果分别为“ 82”和“（）”。

Answer 1

我相信您正在以错误的顺序调用带有参数的函数。附言：我无法发表评论，所以我必须作答

Answer 2

请参阅这篇文章 Error with Sklearn Random Forest Regressor

和解决方案

rgr = regressor.fit（map（lambda x：[x]，X），y）

打印语句确认大小相同时，出现“ ValueError：标签数量与样本数量不匹配”

2 个答案: