在Pandas层次化索引中应用函数

huangapple
go评论77阅读模式
英文:

apply function in pandas hierarchical index

问题

  1.    points  tiger  lion     bear
  2. 0     425     72  1878  1.40000
  3. 1     441   -211 -5238 -4.00000
  4. 2    1048     47 -1266  2.90000
英文:

I have a pandas dataframe as below.

  1. df = pd.DataFrame({'team' : ['A', 'B', 'A', 'B', 'A', 'B'],
  2.                    'tiger' : [87, 159, 351, 140, 72, 119],
  3.                    'lion' : [1843, 3721, 6905, 1667, 2865, 1599],
  4.                    'bear' : [1.9, 3.3, 6.3, 2.3, 1.2, 4.1],
  5.                    'points' : [425, 425, 441, 441, 1048, 1048]})
  7. grouped = df.groupby(['points', 'team'])[['tiger', 'lion', 'bear']].median()
  9. print(grouped)
  11.                 tiger       lion    bear
  12. points team                             
  13. 425    A     87.00000 1843.00000 1.90000
  14.        B    159.00000 3721.00000 3.30000
  15. 441    A    351.00000 6905.00000 6.30000
  16.        B    140.00000 1667.00000 2.30000
  17. 1048   A     72.00000 2865.00000 1.20000
  18.        B    119.00000 1599.00000 4.10000

I would like to take the difference between teams A and B for each of the animal (tiger, lion, bear) and points levels. So the difference between team A (87) and B (159) within points 425 and tiger. I'm not sure how to do this with an hierarchical index. It would look something like below. Thanks.

  1.    points  tiger  lion     bear
  2. 0     425     72  1878  1.40000
  3. 1     441   -211 -5238 -4.00000
  4. 2    1048     47 -1266  2.90000
  6. </details>
  9. # 答案1
  10. **得分**: 1
  12. 你可以使用 [`swaplevel`](https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.swaplevel.html) 和切片：
  14. ```python
  15. grouped = (df.groupby(['points', 'team'])[['tiger', 'lion', 'bear']].median()
  16.             .swaplevel()
  17.           )
  19. grouped.loc['A'] - grouped.loc['B']

或者使用 xs：

  1. grouped = df.groupby(['points', 'team'])[['tiger', 'lion', 'bear']].median()
  3. grouped.xs('A', level='team') - grouped.xs('B', level='team')

输出结果：

  1.         tiger    lion  bear
  2. points                     
  3. 425     -72.0 -1878.0  -1.4
  4. 441     211.0  5238.0   4.0
  5. 1048    -47.0  1266.0  -2.9
英文:

You can swaplevel and slice:

  1. grouped = (df.groupby([&#39;points&#39;, &#39;team&#39;])[[&#39;tiger&#39;, &#39;lion&#39;, &#39;bear&#39;]].median()
  2.             .swaplevel()
  3.           )
  5. grouped.loc[&#39;A&#39;]-grouped.loc[&#39;B&#39;]

Or use xs:

  1. grouped = df.groupby([&#39;points&#39;, &#39;team&#39;])[[&#39;tiger&#39;, &#39;lion&#39;, &#39;bear&#39;]].median()
  3. grouped.xs(&#39;A&#39;, level=&#39;team&#39;)-grouped.xs(&#39;B&#39;, level=&#39;team&#39;)

Output:

  1.         tiger    lion  bear
  2. points                     
  3. 425     -72.0 -1878.0  -1.4
  4. 441     211.0  5238.0   4.0
  5. 1048    -47.0  1266.0  -2.9

答案2

得分: 0

grouped.groupby(level=0).apply(lambda dd:dd.diff().tail(1)).droplevel([1,2])

out

  1.        老虎    狮子   熊

points
425 72.0 1878.0 1.4
441 -211.0 -5238.0 -4.0
1048 47.0 -1266.0 2.9

英文: 
  1. grouped.groupby(level=0).apply(lambda dd:dd.diff().tail(1)).droplevel([1,2])

out

  1.         tiger    lion  bear
  2. points                     
  3. 425      72.0  1878.0   1.4
  4. 441    -211.0 -5238.0  -4.0
  5. 1048     47.0 -1266.0   2.9

huangapple
  • 本文由 发表于 2023年2月7日 04:02:07
  • 转载请务必保留本文链接：https://go.coder-hub.com/75366021.html
  • group-by
  • indexing
  • pandas
匿名

发表评论

匿名网友

:?: :razz: :sad: :evil: :!: :smile: :oops: :grin: :eek: :shock: :???: :cool: :lol: :mad: :twisted: :roll: :wink: :idea: :arrow: :neutral: :cry: :mrgreen:

确定