Multiple Keepfilters

Thanks, mwegener.  This one is hard to sample.  It's extremely involved.  It involves the following measures (see code at bottom).

Essentially, the ALL() filters when measures get embedded in other measures sometimes dont tack all the way through because I've got a lot of differently-sized tables.  Perfect for KEEPFILTERS(), right?

Ideally, though, instead of   SUMX ( KEEPFILTERS ( VALUES ( //TABLE[COLUMN] to keep filters on// ) ), //MEASURE// )
I could engineer SUMX ( KEEPFILTERS ( // multiple tables that could touch the measure where I need the filters kept // ) ), //MEASURE ).

This is not a critical one at all.     Just trying to figure out advanced KEEPFILTERS ().  Thank you.

Best,

Charlie

\v2_FTE\1 = 
VAR Step1 =
    CALCULATETABLE (
        ADDCOLUMNS (
            SUMMARIZE ( all_articles_ever, DimDate[Year], all_articles_ever[user_id] ),
            "FirstDate", FIRSTDATE ( all_articles_ever[date] ),
            "LastDate", LASTDATE ( all_articles_ever[date] ),
            "^Days_Granular", [^Days_Granular],
            "^Days_alt", [^Days_alt],
            // This the same measure as ^Days_Granual but uses the date table dimdate to determine the available days in a period ---- COUNTROWS(DATESBETWEEN(DimDate[date],FIRSTDATE(DimDate[date]),LASTDATE(DimDate[date])))
            "DailyLocAvg-Articles", [__DailyLocAvg_Art],
            "#Art", [**#Posts_Articles_Patch_AllStaff]
        ),
        all_articles_ever[Lkup_RoleID] IN {4, 5},
        all_articles_ever[type] = "article",
        // all_articles_ever[user_id] <> "23215973" || all_articles_ever[user_id] <> "23083459" || all_articles_ever[user_id] <> "23586758", //This doesnt work. 
         NOT (all_articles_ever[user_id] IN {"23215973", "23083459",  "23586758"}),  
        // could also use && but this seems faster
        ALL ( 'JSON 20181207' )
    ) -- Step 2 - Add Column that qualifies FTEs as needing to have produced >2.5 articles and worked for 32+ days ....and then gives this a haircut on FTEs to the extent their days per the dates on the article are less the dates in the period
VAR Step2 =
    ADDCOLUMNS (
        Step1,
        "@FTE Alt",
        VAR _FTELocalEditManual =
            AND ( [DailyLocAvg-Articles] >= 2.5, [^Days_Granular] > 32 )
        VAR _PctYr360 =
            ROUND ( MIN ( 1, [^Days_Granular] / MIN ( 360, [^Days_alt] ) ), 1 )
        RETURN
            _FTELocalEditManual * _PctYr360
    )
VAR Step3 =
    SUMX ( Step2, [@FTE Alt] )
RETURN
    Step3

\v2_FTE\3 "ALL"_#ArtDstnct_Ed = CALCULATE(
    [^#Articles],
    ALL('JSON 20181207'[wbs_code] ) 
    ) //keep as all arts since table restricts to patch at source)   Dont use deduped since sumn of WBS level (deduped) will be more than the deduped author level (same article can be unique to multiple patches but will be only 1 to author)

\V2_FTE\4 (Req Ed_WBS Filter) = ([^#Articles])/[\v2_FTE\3 "ALL"_#ArtDstnct_Ed]*[\v2_FTE\1]   //keep as articles distinct since that's the pro rata that matters and table restricts to eds earlier
//WBS level is deduped at that level....but author level dedupe means few stories at that level

\v2_FTE\ FINAL = //seems to work.   needs roles 4, 5 selected
SUMX(
    KEEPFILTERS(VALUES(p_user_dump[first_last_name])),
    [\V2_FTE\4 (Req Ed_WBS Filter)]
)