Creating Ranking by Week Measure on top of Virtual Tables

Hello,

My dashboard tracks volume collections daily. I've been able to use RANKX to create a daily ranking measure without an issue. What I would like to do now is give the current selected week a ranking in comparison with all historic weeks.

I've been able to create this ranking in a table as follows:

*temp = 
VAR table1 = SUMMARIZECOLUMNS(
            ORACLE_DATE_DIMENSION[week_end],
            FILTER(
                ORACLE_DATE_DIMENSION, 
                ORACLE_DATE_DIMENSION[WEEK_END]<=[pm_Selected_Period_Week_Ending]
            ),
            "Claim Volume",
            SUM(CLAIM_VOLUMES_DAILY[CLAIM_VOLUME_ADJUST])
    )

VAR table_new = 

ADDCOLUMNS(
    SUMMARIZE(
        table1,
        ORACLE_DATE_DIMENSION[WEEK_END],
        [Claim Volume]
    ),
    "Rank1", RANKX(table1,[Claim Volume],,DESC)
)

return
table_new

The top 5 from that table looks like this:

Now I'd like to create a measure that references these virtual tables but returns the "Rank1" value based on which "WEEK_END" has been selected. I have tried to use the above variables and the following expression, but DAX does not seem to recognize 'table_new' as a proper variable:

*Week collection rank = 
VAR table1 = SUMMARIZECOLUMNS(
            ORACLE_DATE_DIMENSION[week_end],
            FILTER(
                ORACLE_DATE_DIMENSION, 
                ORACLE_DATE_DIMENSION[WEEK_END]<=[pm_Selected_Period_Week_Ending]
            ),
            "Claim Volume",
            SUM(CLAIM_VOLUMES_DAILY[CLAIM_VOLUME_ADJUST])
    )

VAR table_new = 
CALCULATE(
ADDCOLUMNS(
    SUMMARIZE(
        table1,
        ORACLE_DATE_DIMENSION[WEEK_END],
        [Claim Volume]
    ),
    "Rank1", RANKX(table1,[Claim Volume],,DESC)
))
return
maxx(table_new,[Rank1])

That code returns the error "The expression refers to multiple columns. Multiple columns cannot be converted to a scalar value." The "table_new" gets the red squiggly line, as does "Rank1".

Am I correct that only one virtual table can come through as a variable when used in a measure?

If so, how can I complete this calculation with one virtual table variable (or without variables)?

Sincerely,

Nicolas