DAX Select all except topN

[Anonymous]

Hello all

Is there a way to select all values from a table except the topN? 

Sample data can be found here: sample data onedrive 

I want to know the impact of the top 3 worse employees based on time to complete (In sample file that C D & G). Is there a way to exclude the top 3 out of the calculation? I wrote the following but that doesn't seem to work because i don't know how to select the value from a SUMMARIZE table:

Ontimewithouttop5 = CALCULATE(SUM(table1[completedintime])/COUNT(table1[orderid]), FILTER(TOPN(5,SUMMARIZE(table1, table1[employeename], "AboveAverageAgent", AVERAGE(table1[timetocomplete]))), NO IDEA HOW TO SELECT THE RIGHT COLUMN <> table1[employeename]))

Any assistance would be appriciated. DAX has the preference.

[v-cherch-msft]

Hi @Anonymous

You may get the table with visual lever filter. For example, you may create a rank measure and use it in visual level filter. 

AboveAverageAgent =
CALCULATE (
    SUM ( Table3[timetocomplete] ) / COUNT ( Table3[employeename] ),
    ALLEXCEPT ( Table3, Table3[employeename] )
)

Rank = RANKX(ALL(table3),[AboveAverageAgent],,DESC,Dense)

If it is not your case, please share some data sample and expected output. 

Regards,

Cherie

[Anonymous]

hi @v-cherch-msft,

Thanks for your answer, but it doesnt quiet fit. It works, but not ideal as i still have to filter "rank" to a top 3. 

This file is going to be used for different teams and its not ideal for the end users to manually change the filter in case there are no double rank 1 and 2's.

Let me give some more information, i see that my original post is lacking that. 

I want to calculate the SLA with the following calculation SLA = (SUM(completedintime)/DISTINCTCOUNT(orderid))*100 

The expected result is the SLA of all employees BUT the top 3 based on worse timetocomplete. To see the  impact on the SLA if the Top (or bottom) 3 

In the samplefile the SLA would be 62 (%)

without the top 3 (C,D&G) it would be 67 (%)

[v-cherch-msft · Accepted solution]

Hi @Anonymous

You may check below measure.

Measure =
VAR a =
    SUMMARIZE (
        sampledata,
        sampledata[employeename],
        "AboveAverageAgent", AVERAGE ( sampledata[timetocomplete] )
    )
VAR b =
    ADDCOLUMNS ( a, "rank", RANKX ( a, [AboveAverageAgent] ) )
RETURN
    CALCULATE (
        SUM ( sampledata[completedintime] ) / DISTINCTCOUNT ( sampledata[orderid] ),
        FILTER ( b, [rank] > 3 )
    )

Regards,
Cherie

[Anonymous]

Hi @v-cherch-msft,

Perfect! does the job just how I want it. I didn't know I could use variables in a measure, a new world just opened up for me haha. Going to have some fun with this, thanks!

[PattemManohar]

@Anonymous It will be great if you can provide us sample data and expected output..... You DON'T need to provide actual data...

[Anonymous]

Added it 🙂