正文
Python中的sklearn--KFold与StratifiedKFold
提示：扫一扫查出行【扫一扫了解最新限行尾号】
复制提示
KFold划分数据集的原理：根据n_split直接进行划分
StratifiedKFold划分数据集的原理：划分后的训练集和验证集中类别分布尽量和原数据集一样
#导入相关packages from sklearn.model_selection import KFold
from sklearn.model_selection import StratifiedKFold
1 import numpy as np
2 from sklearn.model_selection import StratifiedKFold
3 X = np.array([[1, 2], [3, 4], [1, 2], [3, 4]])
4 y = np.array([0, 0, 1, 1])
5 skf = StratifiedKFold(n_splits=2)
6 skf.get_n_splits(X, y)
7 >>>2
8 print(skf)
9 StratifiedKFold(n_splits=2, random_state=None, shuffle=False)
10 for train_index, test_index in skf.split(X, y):
11 print("TRAIN:", train_index, "TEST:", test_index)
12 X_train, X_test = X[train_index], X[test_index] #返回数据框的方法 X_train = X.iloc[train_index]
13 y_train, y_test = y[train_index], y[test_index]
14 >>>TRAIN: [1 3] TEST: [0 2] #返回的是索引
15 >>>TRAIN: [0 2] TEST: [1 3]
