服务器之家

服务器之家 > 正文

python 划分数据集为训练集和测试集的方法

时间:2021-04-28 00:34     来源/作者:心雨心辰

sklearn的cross_validation包中含有将数据集按照一定的比例,随机划分为训练集测试集的函数train_test_split

?
1
2
3
from sklearn.cross_validation import train_test_split
#x为数据集的feature熟悉,y为label.
x_train, x_test, y_train, y_test = train_test_split(x, y, test_size = 0.3)

得到的x_train,y_train(x_test,y_test)的index对应的是x,y中被抽取到的序号。

若train_test_split传入的是带有label的数据,则如下代码:

?
1
2
3
from sklearn.cross_validation import train_test_split
#dat为数据集,含有feature和label.
train, test = train_test_split(dat, test_size = 0.3)

train,test含有feature和label的。

自己写了一个函数:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
#X:含label的数据集:分割成训练集和测试集
#test_size:测试集占整个数据集的比例
def trainTestSplit(X,test_size=0.3):
 X_num=X.shape[0]
 train_index=range(X_num)
 test_index=[]
 test_num=int(X_num*test_size)
 for i in range(test_num):
  randomIndex=int(np.random.uniform(0,len(train_index)))
  test_index.append(train_index[randomIndex])
  del train_index[randomIndex]
 #train,test的index是抽取的数据集X的序号
 train=X.ix[train_index]
 test=X.ix[test_index]
 return train,test

以上这篇python 划分数据集为训练集和测试集的方法就是小编分享给大家的全部内容了,希望能给大家一个参考,也希望大家多多支持服务器之家。

原文链接:https://blog.csdn.net/xidianliutingting/article/details/53463033

相关文章

热门资讯

2020微信伤感网名听哭了 让对方看到心疼的伤感网名大全
2020微信伤感网名听哭了 让对方看到心疼的伤感网名大全 2019-12-26
yue是什么意思 网络流行语yue了是什么梗
yue是什么意思 网络流行语yue了是什么梗 2020-10-11
背刺什么意思 网络词语背刺是什么梗
背刺什么意思 网络词语背刺是什么梗 2020-05-22
苹果12mini价格表官网报价 iPhone12mini全版本价格汇总
苹果12mini价格表官网报价 iPhone12mini全版本价格汇总 2020-11-13
Intellij idea2020永久破解,亲测可用!!!
Intellij idea2020永久破解,亲测可用!!! 2020-07-29
返回顶部