How do I get my DAX measure to calculate grouped values? - powerbi

Question

I need help with a DAX measure in my Power BI report. I am just learning about measures, so sorry if this seems like a newbie question.
Here’s my scenario:
The purpose of the report is to show water usage for various locations in a municipality. The data is coming from IOT sensors, with new entries every few minutes or so.
A sample of the data generally looks like this:
|SensorID |Timestamp |Reading
|----------|---------------------|--------
|1 |2017-06-22 12:01 AM |123.45
|1 |2017-06-22 12:15 AM |124.56
|1 |2017-06-22 12:36 AM |128.38
|2 |2017-06-22 02:12 AM |564.75
|2 |2017-06-22 02:43 AM |581.97
I have a simple DAX measure that I use to calculate water usage for each location/sensor, for the current date range selected in the report (via Timeline control):
Usage:= (MAX([Reading]) - MIN([Reading]))
This measure works great when a single location/sensor is selected. However, when I select multiple locations, the calculation is incorrect. It takes the MAX value from ALL sensors, and subtracts the MIN value from ALL sensors - rather than calculating the usage from each location and then summing the usage.
For example, given the data sample above, the correct calculation should be:
Total Usage = (128.38 - 123.45) + (581.97 - 564.75) = 22.15
Instead, it is calculating it this way:
Total Usage = (581.97 - 123.45) = 458.52
How can I get the measure to calculate the usage, grouped by the Sensor ID?
I hope this makes sense.
Thanks!

Answer 1

Try this:
Total Usage:= SUMX( VALUES(MyTable[SensorID]), [Usage])
VALUES(MyTable[SensorID]) function gives a list of unique SensorIDs. SUMX function then goes over that list one by one and calculates your [Usage] measure per SensorID. Then it sums up the results.
An alternative solution:
Total Usage:= SUMX( SUMMARIZE(MyTable, MyTable[SensorID]), [Usage])
It works the same way, only the list of unique sensor ids is returned by SUMMARIZE function instead of VALUES.
Results: