Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!
Hi there. I have a question.
Imagine. A table. 6+ colums.
1. Business Unit
2. Soldto Party
3. Material
4. Price effect in €
5. Volume effect in €
6. FX effect. in €
7. i.e. Region
8. etc...etc..
I can do all sorts of stuff with this, obviously. However, i need to make a waterfall visual.
Typically, you would need something like this, with 2 columns:
1. Effect type label
2. Effect amount
This you can feed to the waterfall, with 1. as your legend item and 2 resulting in a positive or negative result per item. E.g. Start at 0, +50 for volume effect, -20 for price, +30 for FX, so total +60. But of course, there is also still the filter on BU, region, cross filter wishes, etc.
I cant wrap my head around this. I think in the direction of creating a calculated table like the one sketched above (unpivot?) but how can i make it responsive to the said wishes? There must be a better way then unpivoting 400k+ rows with 40+ columns.
Can anyone help me? It doesnt happen often anymore, but i am really stuck
Solved! Go to Solution.
Solved it myself in beforesaid method.
Made in original table a unique key
Used a copy of the original table
Deleted in copy all non-needed columns
Performed unpivot on the three needed columns.
Made relation between the two tables, single direction to new table.
Dragged new table effect name and values to waterfall.
Done 🙂
Does take a lot of memory perhaps this way?
Solved it myself in beforesaid method.
Made in original table a unique key
Used a copy of the original table
Deleted in copy all non-needed columns
Performed unpivot on the three needed columns.
Made relation between the two tables, single direction to new table.
Dragged new table effect name and values to waterfall.
Done 🙂
Does take a lot of memory perhaps this way?
User | Count |
---|---|
125 | |
106 | |
99 | |
63 | |
62 |
User | Count |
---|---|
135 | |
116 | |
101 | |
71 | |
61 |