On change data obtain with DAX

[Anonymous]

I have a table containing state changes for multiple values

time	primary_key	value	lag_value
1	A	100	0
2	B	200	0
3	A	300	100
4	B	400	200
6	A	0	300
6	B	0	400

I can easily calculate the value of the last state with the following formula:

WantedVal =
var vDate = MAX('Table'[time])
Return
CALCULATE(SUM('Table'[value]),FILTER(ALL('Table'[time]),'Table'[time]<= vDate))-CALCULATE(SUM('Table'[lag_value]),FILTER(ALL('Table'[time]),'Table'[time]<= vDate))

However, I'd rather not create all the lags as I have multiple value variables.

I've tried to recreate this measure without the lag, but I fail to make it work in a context of primary_key + time, primary_key or time.
For instance, the following measure works in primary_key + time but not in a context only using time:

no_lag_value = 
VAR vDate = MAX('Table'[time])
RETURN
CALCULATE(SUM('Table'[value]),FILTER(ALL('Table'[time]),'Table'[time]<= vDate)) -  SUMX(  FILTER(ALLSELECTED('Table'),'Table'[group] in VALUES('Table'[group])),CALCULATE(SUM('Table'[value]),'Table'[time]<vDate))

Summarising, is there a way to reproduce WantedVal without using lag_value?

[amitchandak]

@Anonymous ,

new column =
var _1 = maxx(filter(Table1, Table1[time] < earlier(Table1[time]) && Table1[primary_key] =earlier(Table1[primary_key]) ), [time])
return
maxx(filter(Table1, Table1[time] =_1 && Table1[primary_key] =earlier(Table1[primary_key]) ), [value])

[Anonymous]

Hello @amitchandak ,
I don't want a column with the lag (I could add it in the m query). I have around ten values on change, and more can come making it hard to sustain. For that reason, I'm looking for a measure that I can generalise if needed.