如果时间戳关闭，则删除重复

Question

我有一个关于谁在工作、在哪个任务中以及他/她什么时候开始工作的“日志”信息：

 index |    Entrance time   | Name | Last name | Employee_ID  | Task
 --------------------------------------------------------------------
   0   |2000-01-01 00:00:00 | John |  Fischer  |    001       | Maintenance
   1   |2000-01-01 00:04:30 | John |  Fischer  |    001       | Development
   2   |2000-01-01 00:04:30 | Bob  |  Conrad   |    002       | Maintenance
   3   |2000-01-01 00:10:00 | Mary |  Smith    |    003       | Multitasking
   4   |2000-01-01 00:09:30 | John |  Fischer  |    001       | Maintenance
   5   |2000-01-01 00:15:30 | John |  Fischer  |    001       | Maintenance
   6   |2000-01-02 00:04:30 | Bob  |  Conrad   |    002       | Maintenance
   7   |2000-01-02 00:10:00 | Mary |  Smith    |    003       | Multitasking

然后，我想消除重复，如果入口时间差之间的一个我们正在寻找和其余是低于10分钟，如果任务和名称是相同的。因此，由此产生的数据应该是：

 index |    Entrance time   | Name | Last name | Employee_ID  | Task
 --------------------------------------------------------------------
   0   |2000-01-01 00:00:00 | John |  Fischer  |    001       | Maintenance
   1   |2000-01-01 00:04:30 | John |  Fischer  |    001       | Development
   2   |2000-01-01 00:04:30 | Bob  |  Conrad   |    002       | Maintenance
   3   |2000-01-01 00:10:00 | Mary |  Smith    |    003       | Multitasking
   5   |2000-01-01 00:15:30 | John |  Fischer  |    001       | Maintenance
   6   |2000-01-02 00:04:30 | Bob  |  Conrad   |    002       | Maintenance
   7   |2000-01-02 00:10:00 | Mary |  Smith    |    003       | Multitasking

我使用了drop_duplicates(subset="Name“、"Last”、"Task")，但我不知道如何应用时间条件来比较每一行和其余行。

希望你能帮我，谢谢

Answer 1

对于时间差的计算，这可能会对你有所帮助。但是，您还需要根据重复情况应用您的条件。

# Make df sequential in ["Name", "Last name", "Task"]
df.sort_values(["Name", "Last name", "Task"], inplace=True)

# Compute time difference 
temp = df['Entrance time'] - df['Entrance time'].shift()

# converts the difference in terms of minutes (taking into account absolute values)
df['diff_mins'] = temp.abs() /np.timedelta64(1,'m')

产出：

2  2  2000-01-01 00:04:30  Bob   Conrad   2  Maintenance    nan
6  6  2000-01-02 00:04:30  Bob   Conrad   2  Maintenance   1440
1  1  2000-01-01 00:04:30  John  Fischer  1  Development   1440
0  0  2000-01-01 00:00:00  John  Fischer  1  Maintenance      4.5
4  4  2000-01-01 00:09:30  John  Fischer  1  Maintenance      9.5
5  5  2000-01-01 00:15:30  John  Fischer  1  Maintenance      6
3  3  2000-01-01 00:10:00  Mary  Smith    3  Multitasking     5.5
7  7  2000-01-02 00:10:00  Mary  Smith    3  Multitasking  1440