Filter by month on count distinct

[Anonymous]

Hi Everybody

I have a list of project as a source to my report. For each project there is a column which say which platform was used in this specific project, if non the cell is blank. I have counted the number of unique used platforms for these projects, by creating a measure:

Distinct Count of Platform used difference from (Blank)=

VAR __BASELINE_VALUE =

    CALCULATE(

        DISTINCTCOUNT('Investments'[Platform used]),

        ISBLANK('Investments'[Platform used])

    )

VAR __MEASURE_VALUE = DISTINCTCOUNT('Investments'[Platform used])

RETURN

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

This shows me correct result on overall level (all months), however I have to be able to filter this by months.

If I do this using the formula above the result is incorrect, I think that it first filter the projects by month and then counts distinct difference from blanks. What I need to get is number of unique platforms used overall in each month.

Thanks in advance for any kind of help!

[amitchandak]

@Anonymous , Can you share sample data and sample output in table format? Or a sample pbix after removing sensitive data.

With what is wrong and what is expected .

Also, try this return

RETURN
sumx(Values('Investments'[Month Year]),IF(NOT ISBLANK(__MEASURE_VALUE), __MEASURE_VALUE - __BASELINE_VALUE))

[Anonymous]

If I include all data the visual table shows correct info about unique used platforms. However if I filter it by month(for ex August) then platform used(BX2) in other month(February) shows as unique used in August which is not true. I want it to show only platforms used for the first time ever(all data) as unique used in each month.

Unfortunately RETURN that you proposed do not solve this problem.