Why a calculated table can not be ranked with RANKX()? The result is 1 for every rows!

[Yiyi]

Hello friend, 

Could you kindly help me to figure out how to rank a calculated table? 

I have one table as my original table and based on the original table, I have created a new table to count unique values for each media. I have also added a rank for the new table. The DAX is as follows: 

SummarizedTable =
    ADDCOLUMNS(
        SUMMARIZE(
            ALLSELECTED(Data),
            Data[Media],
            "Unique_Item_Count", DISTINCTCOUNT(Data[HeadlineID])
        ),
        "Rank",
        RANKX(ALLSELECTED(Data[Media]), [Unique_Item_Count], , DESC,Dense)
    )

 The result I got is as follows: 

I really appreciate if you could help! Thank you.

[eliasayyy · Accepted solution]

please try 

SummarizedTable = 
VAR _t =
        SUMMARIZE(
            ALLSELECTED('Table'),
            'Table'[Media],
            "Unique_Item_Count", SUM('Table'[Count])
        )
RETURN
ADDCOLUMNS(
    _t,
    "Rank",
    RANKX(_t,[Unique_Item_Count],,DESC,Dense)
)

[Yiyi]

SummarizedTable = VAR _t= SUMMARIZE( ALLSELECTED('Main Data'), 'Main Data'[Media], "Unique_Item_Count", DISTINCTCOUNT('Main Data'[HeadlineID]) ) RETURN RANKX(_t,[Unique_Item_Count]+RAND(),,DESC,Dense)

[Yiyi]

Hello thanks for your help again! I am trying to get the ranks for the running total calculation thus I need the ranks to be distinct number. I found somewhere saying that by using RAND() might help to achieve such result. Do you know how to do it? Thanks a millions time! 

[eliasayyy]

Spoiler

If you are having trouble with dense and having same rank , it is true you can use rand() which will add a decimal to the total value so [unique measure] + rand() should lpwer the change to get same rank

[Yiyi]

works perfectly, you are like a magician 😉 Thanks a lot 😊