Distinctcount and totals per category in visuals

[ddorhout]

Dear PowerBI community,

I am facing an issue with a visualization (stacked column). I am not sure if I need to make a new row…

I would like to use distinct count of my rows within my columns. I think it is explained best by giving an example:

I use something like this kind of dataset:

DummyX	UserID	Department
Yes	1	X
No	1	X
Yes	1	X
No	2	X
No	2	X
Yes	2	X
Yes	3	Y
Yes	3	Y
Yes	3	Y
No	4	Y
Yes	4	Y

I would like to summarize how many people are having a "No" in the first column, compared to the total number of Users per department.

If there is one row value in which there is a "No", then the count function should judge a user to be only counted as a "No".

So a figure (it is supposed to be a stacked column) would look then like the following, stacked on each other.

		1x Yes
2x No		1x No
X		Y

At the moment, by using the distinctcount functionality in my visualization I get a visualization of the following in which some users are counted double if they have both a "Yes" and a "No":

2x Yes		2x Yes
2x No		1x No
X		Y

I hope you could help me out… 🙂

Best,

Doeke

[ddorhout · Accepted solution]

Thank for your reply! I understand that my explanation was not totally clear...

Eventually, I found the solution myself.

I combined two measures in my table:

1.

Count only Yes=

VAR __BASELINE_VALUE =

    CALCULATE(

        DISTINCTCOUNT('Table'[UserID]),

        'Table'[DummyX]

            IN { "No" }

    )

VAR __MEASURE_VALUE = DISTINCTCOUNT('Table'[UserID])

RETURN

    IF(NOT ISBLANK(__MEASURE_VALUE), __MEASURE_VALUE - __BASELINE_VALUE)

2.

Then I also made a count for the number of "No":

Count only No =

VAR __BASELINE_VALUE =

    CALCULATE(

        DISTINCTCOUNT('Table'[UserID]),

        'Table'[DummyX]

            IN { "No" }

    )

RETURN

__BASELINE_VALUE

Anyway, thanks for your help! 🙂

[YukiK]

Glad to hear that! Please consider giving a thumbs up to my reply if that helped 🙂

[YukiK]

I'm not sure if I understand what your final result table looks like, but here is a DAX measure to unique users who have "No" value in the first column.

Unique Users with At Least 1 No =
VAR __Users =
ADDCOLUMNS(
SUMMARIZE(
'Table',
'Table'[UserID]
),
"@MinDummy", CALCULATE( MINX( 'Table', 'Table'[DummyX] ) )
)
VAR __Res =
CALCULATE(
[Total Unique Users],
FILTER( __Users, [@MinDummy] = "No" )
)
RETURN
__Res

If this helps, please give it a thumbs up!