Find value between two tables and dates

[Yggdrasill]

Maybe it's because it's monday or something but I just can't get my head around a what I thought was a simple stuff but here I am, few hours later...

Problem:

I have two tables with no active relationship

First table is my fact table which holds all data related to punched hours

FactTable

Date	Hours	EmployeeId
29.12.2020	8	A
30.12.2020	8	A
31.12.2020	4	A
1.1.2021	4	A
2.1.2021	8	A

My second table has information on salary rates per employee with a given start date like so:

SalaryTable

EmployeeId	SalaryFromDate	Rate
A	1.1.2020	50
A	1.1.2021	100
B	1.1.2009	60

My desired result

Using Dax measure I want to calculate the cost for each hours that is actually punched in and worked.

I'm trying to use this for example

Rate_Hourly =
VAR _salaryfromdate =
    SELECTEDVALUE ( SalaryTable[SalaryTableFromDate] )
RETURN
    CALCULATE (
        SUM ( SalaryTable[hourlySalaryTable] ),
        FILTER (
            FactTable,
            FactTable[date] >= _salaryfromdate
                && FactTable[date] <= TODAY ()
        )
    )

But this just gives me sum of all Hour rates.

How can I make sure I find the correct Hour rate within the date range given from FactTable and compare to SalaryTable?

Kindest

[sevenhills]

I tried this and see if it works for you

No relationships created

Measure

Rate_Hourly =  
var _Rate = MAXX( 
    FILTER( SalaryTable, SalaryTable[EmployeeID] = SELECTEDVALUE(FactTable[EmployeeID])
                     && SalaryTable[SalaryFromDate] <= SELECTEDVALUE(FactTable[Date]))
                     , SalaryTable[Rate]
) 
return _Rate

FYI: I prefer to have a calculated column than measure if it is a large dataset

... 

[Yggdrasill]

@sevenhills Thanks for the prompt reply and taking the time to help out!

I've successfully implemented the DAX code and it works. However it breaks at another granular level when I'm not looking at dates. See how the total does not sum up

Can I somehow make this work without having to use ISINSCOPE() or other similar functions?

Or should I just do as you mentioned and use a column?

Kind regards

[sevenhills]

Please can you provide what is expected output in your top scenario (like a mockup)

Tough to visualize sometimes ... 

[Yggdrasill]

The expected output is to find hourly rate and multiply that number with the amount of hours worked. That should define the cost. However I need to be able to look at cost regardless of if I'm looking at one employee, two or all.

Come to think of it maybe it's easier to let PQ calculate the cost in the fact table? So everytime an hour is registered another column is calculated for the cost behind that hour

[sevenhills]

• DAX I provided the measure is from your original post and based on a date.
• What you are asking is SCD2 - Fact type of example.
• Let me explain, in case you have not noted already as this type.
  • You have two tables
  • One table is fact i.e., containing transactions. in your case, employee worked hours on each date.
  • Second table is SCD i.e., Slowly change dimension. Where you have effective start date and rate for each employee. (and you are missing is effective end date, which is not an issue).
  • What you are looking is to get the rate based on the transaction date!
• You can take a look at these examples and fit to your needs.
  • https://p3adaptive.com/2019/06/how-can-i-get-a-lookup-table-from-a-slowly-changing-dimension-scd/
  • https://www.zartis.com/scd-implementation-with-temporal-tables-in-power-bi/

If my analysis of your problem is wrong, then I need more details.

Optionally, you can take a look 

https://community.powerbi.com/t5/Community-Blog/HR-Analytics-Active-Employee-Hire-and-Termination-tr...