Running Monthly Total Calculation or Filter

Hii E_Rye 

​In your data, an ID has a "Life Cycle" that starts at EpisodeStartDate and ends at DeliveryDate (or EpisodeEndDate). A standard running total often ignores the end date, causing IDs to stay in the count forever.

​The Solution: The "Active ID" Pattern

​Step 1: Set up a Disconnected Date Table

​For dynamic filtering to work across a 24-month trend, your X-axis should come from a separate Calendar table that is not directly linked to your Episodes dates. This allows the measure to compare the "Month being viewed" against the "Dates in the row."

​Step 2: Create the "Active ID" Logic

​Create a measure that determines if an ID is active during a given month. Use the logic you described: EpisodeStartDate must be on or before the end of the month, and the ID must not have been "Delivered" before the start of that month.

Active ID Count = 
VAR _ReportMonthStart = MAX('Calendar'[Date])
VAR _ReportMonthEnd = EOMONTH(_ReportMonthStart, 0)

RETURN
CALCULATE(
    DISTINCTCOUNT(Episodes[ID]),
    KEEPFILTERS(
        Episodes[EpisodeStartDate] <= _ReportMonthEnd && 
        (ISBLANK(Episodes[DeliveryDate]) || Episodes[DeliveryDate] >= _ReportMonthStart)
    )
)

Step 3: Create the Running Total

​Now, wrap that "Active" logic into a running total calculation. This will iterate through all months up to the one currently being viewed in your chart.

Running Monthly Total = 
VAR _MaxDate = MAX('Calendar'[Date])

RETURN
CALCULATE(
    [Active ID Count],
    FILTER(
        ALLSELECTED('Calendar'),
        'Calendar'[Date] <= _MaxDate
    )
)

Alternative: Using a Visual-Level Filter

​If you want to keep your existing simple count measure, you can create a "Flag" measure:

1. ​Create a measure: FilterFlag = IF([Active ID Count] > 0, 1, 0).
2. ​Drag this into the Filters on this visual pane for your chart.
3. ​Set the filter to "is 1".

If this solution worked for you, please Mark as Solution! It helps others in the community find this answer more easily. Cheers!