如何解决 sklearn 错误：“发现样本数量不一致的输入变量”？

Question

我在使用 sklearn 70-30 部门时遇到了挑战。我在线收到错误：

X_train, X_test, y_train, y_test = train_test_split(X_smote, y_smote, test_size=0.3, stratify=y)

错误是：

Found input variables with inconsistent numbers of samples

上下文

from imblearn.over_sampling import SMOTE
    
sm = SMOTE(k_neighbors = 1)
X = data.drop('cluster',axis=1)
y = data['cluster']
    
X_smote, y_smote= sm.fit_sample(X,y)
    
data_bal = pd.DataFrame(columns=X.columns.values, data=X_smote)
data_bal['cluster']=y_smote
    
from sklearn.model_selection import  train_test_split
X_train, X_test, y_train, y_test = train_test_split(X_smote, y_smote, test_size=0.3, stratify=y)
y_train.value_counts().plot(kind='bar')

编辑

我解决了错误，我只需将stratify=y 放入stratify=y_smote

Answer 1

只是对您的代码行的观察：

X_train, X_test, y_train, y_test = train_test_split(X_smote, y_smote, test_size=0.3, stratify=y)

引发的错误通常是某些输入值的结果，该输入值预期具有与其他输入值一致的特定维度或长度。

检查 X_smote、y_smote 和 y 的长度和/或尺寸，看看它们是否都符合预期。