Forum Discussion
New Measure, search and return
Hello everyone,
Can you please help to tell:
How to create measure to:
Sum Table2[Clicks] for certain Unique [Table1]Campaign Page ID are cointained in Table2[Page] under certain [Table1]Action Week?
e.g. Sum Clicks for all Unique [Table1]Campaign Page ID in "Week10", or Sum Clicks for all Unique IDs in "Update Copy"
Table2:
Thanks.
H
- Anonymous5 years ago
HI Anonymous,
I'd like to suggest you add a new calculated column to table2 to extract the fact page id:
PageID = PATHITEM ( SUBSTITUTE ( [Page], "/", "|" ), PATHLENGTH ( SUBSTITUTE ( [Page], "/", "|" ) ) )Then you can use this field and the date range of the 'week number' to summary records.
Measure = VAR currDate = MAX ( Table[Action Date] ) VAR _list = CALCULATETABLE ( VALUES ( Table[Campaign Page ID] ), FILTER ( ALLSELECTED ( Table ), YEAR ( [Action Date] ) = currDate ), VALUES ( Table[Action Week] ) ) RETURN CALCULATE ( SUM ( Table2[Clicks] ), FILTER ( ALLSELECTED ( Table2 ), [PageID] IN _list && WEEKNUM ( [Date] ) = WEEKNUM ( currDate ) && YEAR ( [Date] ) = YEAR ( currDate ) ) )Regards,
Xiaoxin Sheng
8 Replies
- Greg_Deckler
Community Champion
Anonymous I'm not entirely certain what you are going for exactly but seems like a SUMX over a SUMMARIZE table. Question I have is what constitutes "unique ids"? Is that your Page?
- AnonymousNot applicable
Hi Greg_Deckler ,
Please let me clarify in a more clearer way.
In Table1, [ID] stores value, they are repeated some times, by [Week] or [Type]
the [ID] values contained in [Page] in Table2
Please let me know if there is a way to create Measure, to Sum Impressions for all ID value by Week, or only for one week, such as: WK10.
Repeated ID shall only be Sum by 1 time.
That's say, in WK10, Sum of Impressions of news012808 shall not double.
Appreciated if you can help.
BTW: in fact I asked similar question in this post, but didn't have answer, so I tried to think and ask in another way like this post.
Thanks.
H
- Ashish_Mathur
Super User
Hi,
In an MS Excel file, paste both datasets and show the expected result very clearly, along with formulas