panda为列中的连续值指定累积计数

Question

这是我的数据：

print(n0data)

                          FULL_MPID            DateTime  EquipID  count
Index                                                                  
1      5092761672035390000000000000 2018-11-28 00:36:00     1296      1
2      5092761672035390000000000000 2018-11-28 00:37:00     1634      2
3      5092761672035390000000000000 2018-11-28 13:36:00     1296      3
4      5092761672035390000000000000 2018-11-28 13:38:00     1634      4
5      5092761672035390000000000000 2018-11-29 17:37:00     1290      5
6      5092761672035390000000000000 2018-11-29 17:37:00     1634      6
7      5092761672035390000000000000 2018-11-30 21:23:00     1290      7
8      5092761672035390000000000000 2018-11-30 21:24:00     1634      8
9      5092761672035390000000000000 2018-12-02 09:37:00     1296      9
10     5092761672035390000000000000 2018-12-02 09:39:00     1634     10
11     5092761672035390000000000000 2018-12-02 09:39:00     1634     11
12     5092761672035390000000000000 2018-12-03 11:55:00     1290     12
13     5092761672035390000000000000 2018-12-03 12:02:00     1634     13
14     5092761672035390000000000000 2018-12-06 12:22:00     1290     14
15     5092761672035390000000000000 2018-12-06 12:22:00     1634     15
16     5092761672035390000000000000 2018-12-06 12:22:00     1634     16
17     5092761672035390000000000000 2018-12-06 12:23:00     1634     17
18     5092761672035390000000000000 2018-12-06 12:23:00     1634     18
19     5092761672035390000000000000 2018-12-06 12:23:00     1634     19
20     5092761672035390000000000000 2018-12-06 12:23:00     1634     20
21     5092761672035390000000000000 2018-12-06 12:23:00     1634     21
22     5092761672035390000000000000 2018-12-09 05:51:00     1290     22

因此，我有一个groupBy函数，它使用命令生成以下ecount列：

^{pr2}$

根据时间对数据进行排序，以确定设备何时发生转换。在

Ecount应该是：

当equipeCount的值从一列更改为另一列时，应重置另一列的值。但是，如果EquipID没有改变，比如在索引15-21行期间，EquipID应该继续计数。我还以为这就是groupBy提供的。。。在

Answer 1

您可以在groupby之前使用shift和cumsum技巧：

v = df.EquipID.ne(df.EquipID.shift())
v.groupby(v.cumsum()).cumcount() + 1

Index
1     1
2     1
3     1
4     1
5     1
6     1
7     1
8     1
9     1
10    1
11    2
12    1
13    1
14    1
15    1
16    2
17    3
18    4
19    5
20    6
21    7
22    1
dtype: int64