Previous Value

 Hi edgar 

Please try this measure:

PREVIOUS VALUE = 
VAR _LastV =
    CALCULATE (
        MAX ( 'Table'[VALUE] ),
        FILTER (
            FILTER (
                ALLEXCEPT ( 'Table', 'Table'[NAME] ),
                'Table'[ID] = FIRSTNONBLANK ( 'Table'[ID], "" )
            ),
            'Table'[DATE] < MAX ( 'Table'[DATE] )
        )
    )
RETURN
    IF ( ISBLANK ( _LastV ), 0, _LastV )

The Output would be as below:

Did I answer your question? Mark my post as a solution!

Appreciate your Kudos  !!

Hi,

You may download my PBI file from here.

Hope this helps.

Hi edgar ，

If you can create more than one column, you can use the following approach to add two new columns. This method is easier to understand, first group sort and use the sort order number for matching.

index = 
RANKX (
    FILTER ( 'Table', 'Table'[ID] = EARLIER ( 'Table'[ID] ) ),
    'Table'[DATE],
    ,
    ASC,
    DENSE
)

pre = 
IF (
    CALCULATE (
        MAX ( 'Table'[VALUE] ),
        FILTER (
            'Table',
            'Table'[ID] = EARLIER ( 'Table'[ID] )
                && 'Table'[index]
                    = EARLIER ( 'Table'[index] ) - 1
        )
    )
        = BLANK (),
    "0",
    CALCULATE (
        MAX ( 'Table'[VALUE] ),
        FILTER (
            'Table',
            'Table'[ID] = EARLIER ( 'Table'[ID] )
                && 'Table'[index]
                    = EARLIER ( 'Table'[index] ) - 1
        )
    )
)

If you can only create one column, you can use the following dax.

pre2 = 
VAR _max =
    MAXX (
        FILTER ( 'table', [ID] = EARLIER ( [ID] ) && [DATE] < EARLIER ( [DATE] ) ),
        [DATE]
    )
RETURN
    IF (
        MAXX ( FILTER ( 'table', [ID] = EARLIER ( [ID] ) && [DATE] = _max ), [VALUE] )
            = BLANK (),
        0,
        0
            + MAXX ( FILTER ( 'table', [ID] = EARLIER ( [ID] ) && [DATE] = _max ), [VALUE] )

Final result:

Don't forget to give thumbs up and accept this as a solution if it helped you!!!

Best Regards

Lucien