Forum Discussion
Solved
DAX measure with left outer join, multiple to multiple
I have one table in data model with one row per employee, their current end previous manager, and their current and previous output:
| EmployeeID | Current Manager | Current Output | Previous Manager | Previous Output |
| 1 | A | 10 | B | 5 |
| 2 | A | 20 | A | 15 |
| 3 | B | 30 | A | 40 |
| 4 | B | 5 | B | 10 |
In the reports view, I show the sum of current and previous output by current and previous manager respectively:
| Current Manager | Current Output |
| A | 30 |
| B | 35 |
And:
| Previous Manager | Previous Output |
| A | 55 |
| B | 15 |
I would like to left join the 'current output' table with the 'previous output' table on Current Manager = Previous Manager so I can calculate the output delta. What I want is this:
| Current Manager | Current Output | Previous Output | Delta Output |
| A | 30 | 55 | -25 |
| B | 35 | 15 | 20 |
I have tried writing a measure to give the answer, but can't figure out how to pass the values of the delta:
Create Delta Table =
VAR x2=GENERATEALL(
SUMMARIZE('EmployeeTable', EmployeeTable'[Current Manager],
"Sum Current Output", CALCULATE (SUM('EmployeeTable'[Current Output])))
,SUMMARIZE('EmployeeTable', 'EmployeeTable'[het_name],
"Sum Previous Output", CALCULATE (SUM('EmployeeTable'[Previous Output])))
)
RETURN
x2
You can create measures like
Current Output Measure = SUM( 'Table'[Current Output]) Previous Output Measure = CALCULATE( SUM( 'Table'[Previous Output]), REMOVEFILTERS( 'Table'[Current Manager]), TREATAS( VALUES( 'Table'[Current Manager]), 'Table'[Previous Manager]) ) Delta = [Current Output Measure] - [Previous Output Measure]
2 Replies
You can create measures like
Current Output Measure = SUM( 'Table'[Current Output]) Previous Output Measure = CALCULATE( SUM( 'Table'[Previous Output]), REMOVEFILTERS( 'Table'[Current Manager]), TREATAS( VALUES( 'Table'[Current Manager]), 'Table'[Previous Manager]) ) Delta = [Current Output Measure] - [Previous Output Measure]This works well and is so much simpler than what I was trying to do. Thank you.
