脚本之家,脚本语言编程技术及教程分享平台!
分类导航

Python|VBS|Ruby|Lua|perl|VBA|Golang|PowerShell|Erlang|autoit|Dos|bat|

服务器之家 - 脚本之家 - Python - 浅析pandas随机排列与随机抽样

浅析pandas随机排列与随机抽样

2021-08-28 00:19元小疯 Python

这篇文章主要介绍了pandas随机排列与随机抽样的相关资料,帮助大家更好的利用pandas进行数据分析,感兴趣的朋友可以了解下

随机排列

利用 numpy.random.permutation() 函数,可以返回一个序列的随机排列。将此随机排列作为 take() 函数的参数,通过应用 take() 函数就可实现按此随机排列来调整 Series 对象或 DataFrame 对象各行的顺序。
其示例代码 example1.py 如下:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
import numpy as np
import pandas as pd
#创建DataFrame
df = pd.DataFrame(np.arange(12).reshape(4,3))
print(df)
 0 1 2
0 0 1 2
1 3 4 5
2 6 7 8
3 9 10 11
 
#创建随机排列
order = np.random.permutation(4)
#通过随机排列调整DataFrame各行顺序
newDf = df.take(order)
print(newDf)
 0 1 2
2 6 7 8
3 9 10 11
0 0 1 2
1 3 4 5

随机抽样

随机抽样是指随机从数据中按照一定的行数或者比例抽取数据。随机抽样的函数如下:

numpy.random.randint(start,end,size)

函数中的参数说明如下:

  • start:随机数的开始值;
  • end:随机数的终止值;
  • size:抽样个数。

通过 numpy.random.randint() 函数产生随机抽样的数据,通过应用 take() 函数就可实现随机抽取 Series 对象或 DataFrame 对象中的数据。其示例代码 example2.py 如下

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
import numpy as np
import pandas as pd
#创建DataFrame
df = pd.DataFrame(np.arange(12).reshape(4,3))
print(df)
0 1 2
0 0 1 2
1 3 4 5
2 6 7 8
3 9 10 11
 
#随机抽样
order = np.random.randint(0,len(df),size=3)
#通过随机抽样抽取DataFrame中的行
newDf = df.take(order)
print(newDf)
0 1 2
0 0 1 2
1 3 4 5
1 3 4 5

以上就是详解pandas随机排列与随机抽样的详细内容,更多关于pandas随机排列与随机抽样的资料请关注服务器之家其它相关文章!

原文链接:https://www.cnblogs.com/aitree/p/14309480.html

延伸 · 阅读

精彩推荐