UserID | Channel | Timeslot |
4 | 25 | 08:30 |
4 | 25 | 09:00 |
4 | 25 | 09:30 |
5 | 21 | 08:30 |
5 | 21 | 09:00 |
6 | 20 | 08:30 |
6 | 17 | 09:00 |
7 | 22 | 08:30 |
7 | 22 | 09:00 |
7 | 21 | 09:30 |
Above, what I would like to show is users that have watched a channel once, twice, or three times. The caveat is, this can only be back to back. So if a user watches at 8:00, then goes to another channel at 9:00, then comes back at 9:30 to that channel, then it is only considered 1x for each view. If the user stays on the same channel for multiple timeslots, it can be 2x, 3x.
For example:
User 4 has watched the same channel (25) in three timeslots back to back.
User 5 has watched the same channel (21) two slots back to back
User 6 has watched only 1 channel (20) once, and another (17) once
User 7 has watched channel 22 back to back but has only watched channel 21 once
This may be unclear, but hopefully someone can help!
Hi @mroker
In this case, I think you'll want to try out the multi-row formula tool. First make sure your data is sorted by user then time. Next, use a multi-row formula tool to create a new numeric field (below called New Field), grouping by user, with the following expression:
iif([Channel]=[Row-1:Channel],[Row-1:New Field]+1, 1)
With this, you're basically making a count that increases whenever the same channel continues back to back for each person.
From here, you can use a summarize to get the max count per person per channel, or something.
Hope this helps!
@mroker Look closely at your expression. You may be missing some brackets around field names.