Incorrect Totals when using measure in table

[poojithas]

Hi All,

I am using a measure with the below formula in my table.

When I use this measure in a table, I am getting incorrect totals. The expected total is 1164 but I get 1160.

I tried using a calculated column instead of a calculated measure, but the results are still inacurate.

Could someone please assist on this.

Thank you,

Poojitha

[tackytechtom · Accepted solution]

Hi @poojithas ,

When it comes to non-additive measures (DISTINCTCOUNT is one of them) there is no easy one solution fits all approach.

I recommend you the following post by sqlbi. There is loads to take away from it:

https://www.sqlbi.com/articles/why-power-bi-totals-might-seem-inaccurate/

Hope this helps 🙂

/Tom
https://www.tackytech.blog/
https://www.instagram.com/tackytechtom/

[poojithas · Accepted solution]

This is fixed using the below logic:
=sumx(values(locn),measure2)

[tackytechtom]

Hi @poojithas ,

Here a an example which might give you an idea on why your total differs from the expected outcome:

On the left, a table where each block has a different value. In total there are 4 different values (1, 2, 3, 4), but when grouping by block and using a DISTINCTCOUNT on "Value", we can see that each block might have a different number of distinct values, i.e. California has 3 (1, 2, 3) while Florida has 2 (1, 2). The total itself does not sum over what is shown above in the table above but instead calculates the Measure without applying a filter. This means it total just returns all distinct values, that is 4 (1, 2, 3, 4).

Let me know if this helps 🙂

/Tom
https://www.tackytech.blog/
https://www.instagram.com/tackytechtom/

[poojithas]

@tackytechtom   - Thank you for the details provided on why the totals are differing.

Would like to check if there is a way we can show the expected totals and does not return the distinct values.

[tackytechtom · Accepted solution]

Hi @poojithas ,

When it comes to non-additive measures (DISTINCTCOUNT is one of them) there is no easy one solution fits all approach.

I recommend you the following post by sqlbi. There is loads to take away from it:

https://www.sqlbi.com/articles/why-power-bi-totals-might-seem-inaccurate/

Hope this helps 🙂

/Tom
https://www.tackytech.blog/
https://www.instagram.com/tackytechtom/