Grouping / Aggregation

[Matt_P]

I have the data below and was wondering how to calculate the diff column which is previous qty minus current row quantitiy while ensuring ID is the same.  Many thanks!

ID	RecordDate	Accumilated Qty	Diff
5	06/01/2019	4557
5	13/01/2019	4586	29
5	20/01/2019	4609	23
5	27/01/2019	4621	12
5	10/02/2019	4630	9
5	17/02/2019	4646	16
5	24/02/2019	4658	12
5	03/03/2019	4658	0
5	10/03/2019	4658	0
5	17/03/2019	4690	32
5	24/03/2019	4714	24
5	30/03/2019	4741	27

[V-pazhen-msft]

@Matt_P 

Hi,

I suggest you to create an Index column, this can make it very simple with the following formula:

Result =
VAR current_ = [Index]
RETURN
    IF (
        Sheet1[Index] <> 1,
        [Accumilated Qty]
            - CALCULATE (
                MAX ( [Accumilated Qty] ),
                FILTER ( Sheet1, Sheet1[Index] = current_ - 1 ),
                FILTER ( Sheet1, Sheet1[Index] <> 0 )
            ),
        BLANK ()
    )

Best,
Paul

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[Matt_P]

@V-pazhen-msft this looks like exactly what I'm looking for.  Ill try it out when i get the chance and report back/accept as soution.  Many Thanks!

edit.  Actually, this wouldnt work if I had an id of 4 in the same dataset as the id of 5 would it?

[V-pazhen-msft]

@Matt_P 

This formula is created by looking at the the index column, so it does not matter if your ID is 4 or 5. 
Just leave a message for any updates if needed.

Best, 

[Greg_Deckler]

See my article on Mean Time Between Failure (MTBF) which uses EARLIER: http://community.powerbi.com/t5/Community-Blog/Mean-Time-Between-Failure-MTBF-and-Power-BI/ba-p/3395...

[Matt_P]

Thanks, looked at EARLIER but im finding it difficult to apply... hence the post. 😁