Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!
I hope this problem is not too complex to be covered here.
Imagine you've got an event table like this:
deviceID | timestamp | Feature 1 | Feature 2 | Feature 3 |
D3 | 4/13/2021 | 276 | 130 | 89 |
D14 | 4/24/2021 | 109 | 448 | 240 |
D13 | 4/26/2021 | 79 | 216 | 272 |
There are several calculated features derived from the given features:
(F1+F2)/2 | F1-F2 | F1+F3 | |||||
deviceID | timestamp | Feature 1 | Feature 2 | Feature 3 | Measure 1 | Measure 2 | Measure 3 |
D3 | 4/13/2021 | 276 | 130 | 89 | 203 | 146 | 365 |
D14 | 4/24/2021 | 109 | 448 | 240 | 278,5 | -339 | 349 |
D13 | 4/26/2021 | 79 | 216 | 272 | 147,5 | -137 | 351 |
I now want to apply a threshold check that proves if Measure 1..3 are above certain thresholds that are stored in a separate table:
Effect | Calculation | Property | Threshold1 | Threshold2 | Threshold3 |
E1 | (Feature1+Feature 2)/2 | 90 | 400 | 450 | 480 |
E2 | Feature1-Feature 3 | 140 | 20 | 50 | 100 |
E3 | Feature1+Feature 2 | 90 | 700 | 800 | 850 |
The target is to have one final table that shows which deviceId has exceed which threshold within the last 24h.
I put the pbi dataset here: https://easyupload.io/a4dgtg
I searched if there is already a solution out there but didn't find. What is your suggesestion? How would approach that topic?
Hi @FilipK ,
There's no data in the shared pbix file. And I doubt that you can put different measures in same column.
Best Regards,
Jay
User | Count |
---|---|
139 | |
113 | |
103 | |
73 | |
63 |
User | Count |
---|---|
136 | |
125 | |
107 | |
70 | |
61 |