Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!
Hi All,
Im not sure how to display that kind of information in power bi and if powerbi have relevent charts to that kind of information.
I have a data of users and 5 actions they can do in any order they want with the datetime they actually did that
it looks like
User | ActionA | ActionB | ActionC | Action D
111 26/8/18 25/8/18 22/8/18 21/8/18
112 21/8/18 22/8/18 24/8/18 23/8/18
...
I want to create a funnel chart that will show me the percents of users go from A to B to C to D
and from
A to B to D to C
A > C > B > D
and so on..
anyone have good advice?
or some other idea how i can display these usfull informaiton?
Google Analytics have butiful chart (I think it call snake chart) exactly to that kind of information but i want to create it in my organization dashboards
Thanks
Solved! Go to Solution.
Hi @Drors,
First, unpivot source table in Query Editor mode. And rename columns.
In data view, add calculated columns in data table.
Rank = RANKX ( FILTER ( Dataset1, Dataset1[User] = EARLIER ( Dataset1[User] ) ), Dataset1[Date], , ASC, DENSE ) Action Order = RIGHT ( CALCULATE ( LASTNONBLANK ( Dataset1[Action Type], 1 ), FILTER ( ALLEXCEPT ( Dataset1, Dataset1[User] ), [Rank] = 1 ) ), 1 ) & "-" & RIGHT ( CALCULATE ( LASTNONBLANK ( Dataset1[Action Type], 1 ), FILTER ( ALLEXCEPT ( Dataset1, Dataset1[User] ), [Rank] = 2 ) ), 1 ) & "-" & RIGHT ( CALCULATE ( LASTNONBLANK ( Dataset1[Action Type], 1 ), FILTER ( ALLEXCEPT ( Dataset1, Dataset1[User] ), [Rank] = 3 ) ), 1 ) & "-" & RIGHT ( CALCULATE ( LASTNONBLANK ( Dataset1[Action Type], 1 ), FILTER ( ALLEXCEPT ( Dataset1, Dataset1[User] ), [Rank] = 4 ) ), 1 )
New a calculated table.
Table_1 = ADDCOLUMNS ( FILTER ( CROSSJOIN ( CROSSJOIN ( CROSSJOIN ( VALUES ( Dataset1[Action Type] ), SELECTCOLUMNS ( VALUES ( Dataset1[Action Type] ), "ActionType2", [Action Type] ) ), SELECTCOLUMNS ( VALUES ( Dataset1[Action Type] ), "ActionType3", [Action Type] ) ), SELECTCOLUMNS ( VALUES ( Dataset1[Action Type] ), "ActionType4", [Action Type] ) ), [Action Type] <> [ActionType2] && [Action Type] <> [ActionType3] && [Action Type] <> [ActionType4] && [ActionType2] <> [ActionType3] && [ActionType2] <> [ActionType4] && [ActionType3] <> [ActionType4] ), "Action Order", RIGHT ( [Action Type], 1 ) & "-" & RIGHT ( [ActionType2], 1 ) & "-" & RIGHT ( [ActionType3], 1 ) & "-" & RIGHT ( [ActionType4], 1 ) )
Establish a one to many relationship between above two tables.
Place 'Table_1'[Action Order] and 'Dataset1'[User] into Funnel chart.
Best regards,
Yuliana Gu
Hi @Drors,
First, unpivot source table in Query Editor mode. And rename columns.
In data view, add calculated columns in data table.
Rank = RANKX ( FILTER ( Dataset1, Dataset1[User] = EARLIER ( Dataset1[User] ) ), Dataset1[Date], , ASC, DENSE ) Action Order = RIGHT ( CALCULATE ( LASTNONBLANK ( Dataset1[Action Type], 1 ), FILTER ( ALLEXCEPT ( Dataset1, Dataset1[User] ), [Rank] = 1 ) ), 1 ) & "-" & RIGHT ( CALCULATE ( LASTNONBLANK ( Dataset1[Action Type], 1 ), FILTER ( ALLEXCEPT ( Dataset1, Dataset1[User] ), [Rank] = 2 ) ), 1 ) & "-" & RIGHT ( CALCULATE ( LASTNONBLANK ( Dataset1[Action Type], 1 ), FILTER ( ALLEXCEPT ( Dataset1, Dataset1[User] ), [Rank] = 3 ) ), 1 ) & "-" & RIGHT ( CALCULATE ( LASTNONBLANK ( Dataset1[Action Type], 1 ), FILTER ( ALLEXCEPT ( Dataset1, Dataset1[User] ), [Rank] = 4 ) ), 1 )
New a calculated table.
Table_1 = ADDCOLUMNS ( FILTER ( CROSSJOIN ( CROSSJOIN ( CROSSJOIN ( VALUES ( Dataset1[Action Type] ), SELECTCOLUMNS ( VALUES ( Dataset1[Action Type] ), "ActionType2", [Action Type] ) ), SELECTCOLUMNS ( VALUES ( Dataset1[Action Type] ), "ActionType3", [Action Type] ) ), SELECTCOLUMNS ( VALUES ( Dataset1[Action Type] ), "ActionType4", [Action Type] ) ), [Action Type] <> [ActionType2] && [Action Type] <> [ActionType3] && [Action Type] <> [ActionType4] && [ActionType2] <> [ActionType3] && [ActionType2] <> [ActionType4] && [ActionType3] <> [ActionType4] ), "Action Order", RIGHT ( [Action Type], 1 ) & "-" & RIGHT ( [ActionType2], 1 ) & "-" & RIGHT ( [ActionType3], 1 ) & "-" & RIGHT ( [ActionType4], 1 ) )
Establish a one to many relationship between above two tables.
Place 'Table_1'[Action Order] and 'Dataset1'[User] into Funnel chart.
Best regards,
Yuliana Gu
Wow its looks amaizing, thank you.
I have another question about that, If some of my clients just made 3 actions out of 4, for example
A>B>C (didnt do D)
and I want to see the discrepencies ( in order to know which action is more complicate and the users dont want to do it)
what do you recommend me to do?
Hi @Drors,
It looks like this is a new requirement different from the original post, right? If so, I would suggest you create a new thread on forum so that more community members can see it and provide advice. Please remember to post dummy data and desired result.
By the way, if my above advice is helpful to your scenario, would you please kindly mark it as an answer?
Thanks,
Yuliana Gu
User | Count |
---|---|
139 | |
113 | |
103 | |
73 | |
63 |
User | Count |
---|---|
136 | |
125 | |
107 | |
70 | |
61 |