Calculate moving average of a measure

[springday]

Hi,

We have a table which calculates a measure per day based on the value of a few columns. Now we want to smooth out the measure  to show the trend over time. We tried to use the moving average formula and it throws error with aggregation function as only the table columns can be used there. 

Is this even possible to do in power BI desktop? If not, we'll have to export the data out and caclulate somewhere else? Any ideas how to do this?

Thanks for any help on this.

-Amyee

[parry2k]

@springday Read this post to get your answer quickly.  If you share some sample data, it will help to get you the solution.

https://community.powerbi.com/t5/Community-Blog/How-to-Get-Your-Question-Answered-Quickly/ba-p/38490

[springday]

Here's the sample table.

Date                      ID        Level     Speed  Status 1/28/2019 100 2380 34 1 1/28/2019 200 2198 29 0 1/28/2019 200 2198 35 1 1/28/2019 300 2115 46 0 1/28/2019 400 2396 22 1 1/28/2019 500 1438 12 1

dci = level/ ( median speed when status is 1 + median speed when status is 0)

[v-lionel-msft]

Hi @springday ,

Maybe you can try this measure:

dci = 
VAR x =
CALCULATE(
    MEDIAN(Sheet1[Speed]),
    FILTER(
        Sheet1,
        Sheet1[Status] = 1
    )
)
VAR y = 
CALCULATE(
    MEDIAN(Sheet1[Speed]),
    FILTER(
        Sheet1,
        Sheet1[Status] = 0
    )
)
VAR z =
CALCULATE(
    x+y,
    ALL(Sheet1[Level])
)
RETURN
DIVIDE(
    MAX(Sheet1[Level]),
    z
)

Best regards,
Lionel Chen

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[springday]

Hmm..., it doesn't seem to sovle the moving average problem. Back to my original question, does Power Bi support moving average of a measure?