Pandas Group By 和 Sum 每 N 行 [英] Pandas Group By and Sum Every N rows

问题描述

我有时间序列数据，我想分组并计算每 3 行的总和.似乎是一项简单的任务，但我无法弄清楚.感谢您的帮助.

I have time series data and I want to group by and calculate the sum every 3 rows. Seems like a straightforward task but I'm not able to figure it out. I would appreciate your help.

以下是数据:

df = 
                      AE_NAME  ANSWERED_CALL
    DATE                                    
    2018-10-08  Alec Sochacki            4.0
    2018-10-09  Alec Sochacki            4.0
    2018-10-10  Alec Sochacki            7.0
    2018-10-11  Alec Sochacki            5.0
    2018-10-12  Alec Sochacki            3.0
    2018-10-15  Alec Sochacki            4.0
    2018-10-16  Alec Sochacki            3.0
    2018-10-17  Alec Sochacki            8.0
    2018-10-18  Alec Sochacki            5.0
    2018-10-19  Alec Sochacki            7.0

Column Date 是一个索引列.

Column Date is an index column.

我希望输出如下:

                  AE_NAME  ANSWERED_CALL
DATE                                    
2018-10-08  Alec Sochacki            0 # It's ok to omit the first row
2018-10-10  Alec Sochacki            15
2018-10-15  Alec Sochacki            12
2018-10-18  Alec Sochacki            16
2018-10-19  Alec Sochacki            7 

非常感谢.

推荐答案

首先将您的索引提升为一个系列.然后将 groupby + agg 与字典一起使用:

First elevate your index to a series. Then use groupby + agg with a dictionary:

df = df.reset_index()

d = {'DATE': 'last', 'AE_NAME': 'last', 'ANSWERED_CALL': 'sum'}

res = df.groupby(df.index // 3).agg(d)

print(res)

        DATE       AE_NAME  ANSWERED_CALL
0 2018-10-10  AlecSochacki           15.0
1 2018-10-15  AlecSochacki           12.0
2 2018-10-18  AlecSochacki           16.0
3 2018-10-19  AlecSochacki            7.0

您对第一行的逻辑有些不清楚，因此您可能需要再进行一两次操作.

You've got some unclear logic for the first row, so you may need one or two more operations.

这篇关于Pandas Group By 和 Sum 每 N 行的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！