Using a Multiple filters and a Measure as the Source and Destinations for a Sankey Visual

[SO]

I have data that looks simplified looks something like this 

PersonID		Year		Report		Score		Level
123		20192020		1M		65		L2
134		20192020		1M		55		L1
123		20192020		1M		67		L2
145		20192020		1M		95		L4
123		20202021		1M		75		L3
134		20202021		1M		55		L1
134		20202021		1M		45		R
145		20202021		1M		78		L3

Sankey Mockup

I'd like to produce a Sankey Chart that looks something like above.  

The best I can do is something like this:  

I've tried every combination of source, destination, label, etc. without success.  

I then tried to create to measures so that I could filter out the year and use one measure for the Destination and another for the Destination Labels.

#Average 20192020 =

VAR __BASELINE_VALUE =

    CALCULATE(

        AVERAGE('Table '[Score]),

        'Table'[Year]

            IN { "20192020" }

    )

RETURN

    __BASELINE_VALUE

and 

#Average 20202021 =

VAR __BASELINE_VALUE =

    CALCULATE(

        AVERAGE('Table '[Score]),

        'Table'[Year]

            IN { "20202021" }

    )

RETURN

    __BASELINE_VALUE

I can pull these measures into a table and use them, but they will not drag into any of the Sankey fields.

I'm obviously missing something.  I would like to do this in DAX without going to data tables. 

Thoughts on how to proceed?   Curious ?

SO 

[AllisonKennedy · Accepted solution]

@SO  What is the context of this data? What is it you want to communicate? 

I don't fully understand, but have tried pivoting your data so that you have one column for source and one column for destination, I'm just not sure what to use for the values in these columns (I used Level and set to don't aggregate), but not every person ID has data for both years... 

See file attached below signature.

[AllisonKennedy · Accepted solution]

@SO  What is the context of this data? What is it you want to communicate? 

I don't fully understand, but have tried pivoting your data so that you have one column for source and one column for destination, I'm just not sure what to use for the values in these columns (I used Level and set to don't aggregate), but not every person ID has data for both years... 

See file attached below signature.

[SO · Accepted solution]

Thanks very much Allison.

Each person has 6 scores for each reporting period and there are 6 reporting periods a year and there are 100K entrys per report. Pivoting the table is not an option for me so I was trying to find a way to use a measure to sort this out. We also use AnalysisServices.

You asked a great question, what am I trying to show? I'm trying to find people that have moved from one level to another over time. It appears that most remain at the same level, but some move up, and others down. I'd really like to have the pattern continue over several reports or years, but one step at a time. :-).

Thanks for attempt and much appreciated. I was able to generate my 'ideal' image by creating a small data set and doing what you did. Very intensive to do for each reporting period. Any more thoughts or ideas? I've watched many videos and read many forums and tried many measure codes, but I cannot get it to work for me.