Filtering Issue doughnut Chart - TRICKY

[Anonymous]

Hi Experts

I am having trouble in getting the correct Yes and No values when selecting upward and or downward from the slicer within the attached pbix file. 

Here is the problem when i select from the slicer upward trending, i should only see product (72203976s) which has a value of 416. the doungnut chart below refelcts the correct values in the centre circle 416, but the count of Yes and No (is wrong) it shows No as 404. Which is 14 less this might be assoicated with Yes (missing element from visual).

the measure i have used are in table _reportability table. What i am trying to do is calculate the number of data points that are either Yes or No or Both and those that are upward trending. 

pbix file https://www.dropbox.com/s/5rhg8m3dj9zb6bx/Sample__Test..v1.0.pbix?dl=0

[v-piga-msft]

Hi @Anonymous ,

Firstly, do you only have the incorrect value when selected value is Upward?

If yes, please check if the filter condition in the UP measure is correct as I'm not clear about your logic.

In addition, by my research on your pbix, if we highlight the 404 , it will show the right value like below. 

Last but mot least, please note that do not share your actual data here, always do mask the sensitive data before uploading.

Best  Regards,

Cherry

[Anonymous · Accepted solution]

HI CHerry this works

DR = 
VAR __table = SUMMARIZE(ALL(PMS_PARTS),[Product])
VAR __table1 = ADDCOLUMNS(__table,
                 "__p-value",CALCULATE([p-value],ALL(PMS_COMPLAINT[PMSReportability])),
                        "__TAU",CALCULATE([TAU],ALL(PMS_COMPLAINT[PMSReportability])),
                                "__trend",CALCULATE([__trend],ALL(PMS_COMPLAINT[PMSReportability])))
VAR __table2 = FILTER(__table1,[__trend] = "Down")
VAR __table3 = GENERATE(SELECTCOLUMNS(__table2,"__Product",[Product]),RELATEDTABLE(PMS_COMPLAINT))
RETURN
COUNTAX(__table3,[PMSReportability])

[Greg_Deckler]

@Anonymous  - Looks like you got it!

[Anonymous]

thanks Greg

have a great weekend and enjoy the free time before teh next book. Keep us update if the DAX book is geared more towards Data Science Calculations etc..

[Anonymous]

HI Cherry

if you add the following measure into the matrix table you'll the total is correct all the time. Just the Dougnut chart errors...

NO = CALCULATE(COUNTROWS(PMS_COMPLAINT),FILTER(PMS_COMPLAINT,PMS_COMPLAINT[PMSReportability]="No"))

YES = CALCULATE(COUNTROWS(PMS_COMPLAINT),FILTER(PMS_COMPLAINT,PMS_COMPLAINT[PMSReportability]="Yes"))

[Anonymous]

Hi Cherry

Its supposed to be 416 in total 404 as no and 14 as Yes.