Creating unique table without duplicate values

Hi Molin ,

Thanks for reaching out to Microsoft Fabric Community.

I reviewed your scenario and tried to reproduce it using a small sample dataset with similar structure.
With the below sample data:

Calculated Table:

PU_Times = 
ADDCOLUMNS (
    SUMMARIZE (
        FILTER (
            'TransactionDetails',
            'TransactionDetails'[ObjectNumber] = 999000009
        ),
        'TransactionDetails'[GuestCheckId]
    ),
    "Latest_Reference",
    VAR curCheckId = 'TransactionDetails'[GuestCheckId]
    RETURN
        CALCULATE (
            MAX ( 'TransactionDetails'[Reference] ),
            FILTER (
                'TransactionDetails',
                'TransactionDetails'[GuestCheckId] = curCheckId &&
                'TransactionDetails'[ObjectNumber] = 999000009
            ),
            TOPN (
                1,
                FILTER (
                    'TransactionDetails',
                    'TransactionDetails'[GuestCheckId] = curCheckId &&
                    'TransactionDetails'[ObjectNumber] = 999000009
                ),
                'TransactionDetails'[PostingTime],
                DESC
            )
        )
)

The output was:

Let me know if this meets your requirement. If not, feel free to share a small sample of your dataset or a PBIX file (with any sensitive data removed), and I’ll be glad to take a closer look.

Please consider marking this as the accepted solution if it helps, to assist others facing a similar issue.

Thnak you.

Please find the attached .pbix for reference.