How to filter Pie chart with a measure

[selpaqm]

Customer	Score	grade	Q1	Q2
AA	80	gold	1	1
BB	75	silver	1	1
CC	80	gold	1	1
BB	75	silver	2	2
CC	75	silver	2	2
DD	80	gold	2	2
CC	65	bronze	3	3
DD	75	silver	3	3
EE	100	platinum	3	3

I have a table as above. my aim is to compare in 2 pie chart that only customers belongs to compare months.

İ.e If i compare q1 and q2

bb and cc need to be compared because aa is not in q2 and dd is not in q1. and grade need to be  q1 1 silver 1 gold, q2 2 silver

if compare q1 and q3 only cc need to be copared in pie chart.

I have 2 slicers for quarters in my page and can check changes in table but not in pie chart.

As you can see on power bi pi charts shows 3 customer and comparing them for quarters. however, as you can see on table there are only 2 customer in both quarter and I want to compare them in pie charts. So i need a slicer to see new customers and removed customers and customers that active on both quarter. below you can check the test pbix file.

Test.pbix 

[barritown · Accepted solution]

Hi @selpaqm,

I have a super-complicated solution for your problem with 4 additional auxiliary tables.

If you have a small dataset and won't be frightened by the look of this solution, it will do, otherwise - hope someone proposes a cuter solution.

Here you can download the PBIX file - https://www.dropbox.com/s/p038xljt4ejghy0/community-2-pies-v2.pbix?dl=0

Here's how it works:

1. The initial table is the same but without the Q2 column. It will be added later and will be different from yours.

2. We create 2 additional tables with the help of the following [DAX] code:

Augmented Table 1 = 
VAR _tbl = SELECTCOLUMNS ( 'dataset', 
                           "Customer2", [Customer],
                           "Q2", [Q1] )
VAR _hugetbl = CROSSJOIN ( 'dataset', _tbl )
VAR _filter1 = FILTER ( _hugetbl, [Customer] = [Customer2] )
VAR _filter2 = FILTER ( _filter1, [Q2] <> [Q1] )
RETURN _filter2

You'll get something like this:

Customer2 is not needed here, it is used only for filtering earlier but I don't remove it in order not to make code even more complicated.

3. We create Q1 and Q2 tables which we will use in slicers:

Q1 = VALUES ( 'dataset'[Q1] )

Q2 = SELECTCOLUMNS ( VALUES ( 'dataset'[Q1] ), "Q2", [Q1] )

4. We arrange 4 connections between tables in this auxiliary model:

'Q1'[Q1] -> 'Augmented table 1'[Q1]

'Q1'[Q1] -> 'Augmented table 2'[Q2]

'Q2'[Q2] -> 'Augmented table 1'[Q2]

'Q2'[Q2] -> 'Augmented table 2'[Q1]

5. We use 'Augmented table 1' for the first pie chart and 'Augmented table 2' - for the second.

As you can see, the results are the ones you expect to obtain:

Should you need any clarifications, let me know.

[selpaqm]

@barritown thanks for the solution. it is too complicated but it seems it is working. my dataset has 2k rows and it will be fine.

[selpaqm]

@lbendlin thanks for your kind message. However, this is not what I need. I need 2 pie charts with grades. thanks again.

[lbendlin]

you may be overcomplicating things a little. See attached for my version.

Also note that customer C changes grade from q1 to q3.  Not sure how you plan to handle that.