python之如何优雅的划分train_test_validation 数据?

最新推荐文章于 2024-06-27 21:41:21 发布

Regnaiq

最新推荐文章于 2024-06-27 21:41:21 发布

阅读量1.7k

点赞数 4

分类专栏： Python

本文链接：https://blog.csdn.net/yq_forever/article/details/119325158

版权

Python 专栏收录该内容

34 篇文章 3 订阅

订阅专栏

我们知道skilean有一个train_test_split函数可以帮我们划分trian和test, 那么我们就可以通过这个函数间接完成train_test_validation的划分.

from sklearn.model_selection import train_test_split

def train_test_val_split(train_ratio = 0.8,validation_ratio = 0.1,test_ratio = 0.1,random_state):
    # random_state for reproduction
    # shuffle must be 'True'
    [x_train, x_test, y_train, y_test] = train_test_split(
x, y, test_size=validation_ratio+test_ratio, random_state=random_state, shuffle=True)
    [x_val, x_test, y_val, y_test] = train_test_split(
    x_test, y_test, test_size=test_ratio/(test_ratio + validation_ratio), random_state=random_state)
    return x_train,y_train, x_test, y_test, x_val, y_val