选择数据 - random_rows = df.sample(3) - 随机3行

选择随机n条

import numpy as np
import pandas as pd
df = pd.DataFrame(data=[{'A':2,'B':3,}, {'A':9,'B':8,}, {'A':1,'B':2,}, {'A':4,'B':5,}])

random_rows = df.sample(3)

>>> df.sample(3)
   A  B
1  9  8
0  2  3
2  1  2
>>> df.sample(3)
   A  B
2  1  2
1  9  8
0  2  3
>>> df.sample(3)
   A  B
1  9  8
2  1  2
0  2  3
>>> df.sample(3)
   A  B
1  9  8
3  4  5
0  2  3
>>> df.sample(3)
   A  B
3  4  5
0  2  3
1  9  8
>>> df.sample(3)
   A  B
2  1  2
1  9  8
0  2  3
>>> df.sample(3)
   A  B
2  1  2
1  9  8
3  4  5
>>> df.sample(3)
   A  B
0  2  3
3  4  5
2  1  2
>>> df.sample(3)
   A  B
0  2  3
1  9  8
2  1  2
>>> df.sample(3)
   A  B
0  2  3
3  4  5
1  9  8

可以看见不仅仅随机出现，而且还打乱顺序