Forum Discussion
calculate difference between time stamps for unique ID
Very much obliged for this.
However, I notice that I get a number of columns with errors or blanks once I have pivoted. Is this because there are some duplications where an individual has register 'ACCP' twice for the same ID?
DOH! I didnt' noticed the duplicates in your data. Can you copy/past some sample data (instead of the picture) so we can quickly dump it into Power BI and play with the manipulation... ?
FOrrest
- DR9 years agoNew Member
score time score origin metric 2017-06-24T20:10:04 005056B614ED1ED796A191BDD043A0DA ACCP 2017-06-24T21:43:38 005056B614ED1ED796A191BDD043A0DA CMPN 2017-06-24T20:10:08 005056B614ED1ED796A191BDD043A0DA ENRT 2017-06-24T20:10:14 005056B614ED1ED796A191BDD043A0DA ONST 2017-06-24T19:52:31 005056B614ED1ED796A191BDD043A0DA RCVD 2017-06-24T21:04:05 005056B614ED1ED796A191BDD043A0DA WIP 2017-06-27T06:56:46 005056B614ED1ED796B988341EDC00DA RCVD 2017-06-27T06:59:29 005056B614ED1ED796B988341EDC00DA RCVD 2017-06-27T14:35:38 005056B614ED1ED796B988341EDC00DA RCVD 2017-06-27T14:40:51 005056B614ED1ED796B988341EDC00DA RCVD 2017-06-27T15:47:25 005056B614ED1ED796B988341EDC00DA RCVD 2017-06-27T15:49:41 005056B614ED1ED796B988341EDC00DA RCVD 2017-06-27T18:02:26 005056B614ED1ED796B988341EDC00DA RCVD 2017-05-15T17:08:17 005056B61C081ED78EB0DC4AEE5500D8 ACCP 2017-05-15T17:09:24 005056B61C081ED78EB0DC4AEE5500D8 ENRT 2017-05-15T17:27:43 005056B61C081ED78EB0DC4AEE5500D8 ONST 2017-05-15T17:05:35 005056B61C081ED78EB0DC4AEE5500D8 RCVD 2017-05-15T18:25:20 005056B61C081ED78EB0DC4AEE5500D8 WIP 2017-05-15T20:23:45 005056B61C081ED78EB2523FFFCBE0D8 ACCP 2017-05-15T20:23:50 005056B61C081ED78EB2523FFFCBE0D8 ENRT 2017-05-16T03:55:53 005056B61C081ED78EB2523FFFCBE0D8 NCMF 2017-05-15T20:24:01 005056B61C081ED78EB2523FFFCBE0D8 ONST 2017-05-15T20:14:47 005056B61C081ED78EB2523FFFCBE0D8 RCVD 2017-05-15T20:24:27 005056B61C081ED78EB2523FFFCBE0D8 WIP 2017-05-16T21:42:33 005056B61C081ED78ECEA29F86DEE0D8 ACCP 2017-05-16T22:45:37 005056B61C081ED78ECEA29F86DEE0D8 CMPN 2017-05-16T21:42:42 005056B61C081ED78ECEA29F86DEE0D8 ENRT 2017-05-16T22:04:09 005056B61C081ED78ECEA29F86DEE0D8 ONST 2017-05-16T21:37:00 005056B61C081ED78ECEA29F86DEE0D8 RCVD 2017-05-16T22:41:28 005056B61C081ED78ECEA29F86DEE0D8 WIP 2017-05-18T02:18:23 005056B61C081ED78EECB9F8A64240D8 ACCP 2017-05-18T03:24:16 005056B61C081ED78EECB9F8A64240D8 CMPN 2017-05-18T02:18:25 005056B61C081ED78EECB9F8A64240D8 ENRT 2017-05-18T02:38:02 005056B61C081ED78EECB9F8A64240D8 ONST 2017-05-18T02:12:24 005056B61C081ED78EECB9F8A64240D8 RCVD 2017-05-18T02:47:31 005056B61C081ED78EECB9F8A64240D8 WIP 2017-05-18T03:59:55 005056B61C081ED78EEDE57F4D9720D8 ACCP 2017-05-18T04:49:52 005056B61C081ED78EEDE57F4D9720D8 CMPN 2017-05-18T04:00:00 005056B61C081ED78EEDE57F4D9720D8 ENRT 2017-05-18T04:02:21 005056B61C081ED78EEDE57F4D9720D8 ONST 2017-05-18T03:48:50 005056B61C081ED78EEDE57F4D9720D8 RCVD 2017-05-18T03:57:48 005056B61C081ED78EEDE57F4D9720D8 RCVD 2017-05-18T04:45:50 005056B61C081ED78EEDE57F4D9720D8 WIP 2017-05-18T08:29:41 005056B61C081ED78EF139996DD220D8 ACCP 2017-05-18T20:57:14 005056B61C081ED78EF139996DD220D8 CMPN 2017-05-18T08:29:44 005056B61C081ED78EF139996DD220D8 ENRT 2017-05-18T08:29:52 005056B61C081ED78EF139996DD220D8 ONST 2017-05-18T07:39:20 005056B61C081ED78EF139996DD220D8 RCVD 2017-05-18T20:46:41 005056B61C081ED78EF139996DD220D8 WIP - fhill9 years ago
Resident Rockstar
You were right about the Errors being duplicates. You'll have to descide how you want to handle that data... Since your values are times, Min is Earliest and Max is Latest. The 'null's are blanks in your original data, so ignore them, or go back to your data origin to resolve?
I think the biggest now is 'What if I want First RCVD but Last CMPN??? I don't have an answer for you on that one, sorry... Hopefully someone else jumps on this and might be able to build something simular with custom columns in DAX or something...?
- v-yulgu-msft9 years ago
Microsoft Employee
Hi DR,
Have you worked it out? If yes, please kindly mark the helpful reply as an answer so that it can benefit more users. If you still have any question, please feel free to ask.
Regards,
Yuliana Gu