Difference of a naked column in a calculated column or filter() of a measure

[Applicable88]

Hello,

Measure 1 =

calculate([TotalSales]),

Filter(

ALL('Calendar'[Date]),

'Calendar'[Date]<= MAX('Calendar'[Date])

))

in this Measure 1 above I need to wrap up Calendar[Date] within a aggregation like MAX and somehow it finds the filter context to calendar date. It is actually inside the filter of a interator. If I just write [CalendarDate] I get only the total sum in every row.

But writing the same as a naked colum in a calculated column, which by itself is already an row by row iterator I get right values:

Column= 

Var _TableVariable= Filter( 'Calendar', 'Calendar' [Date]>'Maintable'[BookingDate] && 'Calendar[Day]= "Monday"))
return
countrows(_TableVariable) 

Here I don't need any aggregator I can just write naked column and when computing every row it knows which date and values to fill in. But in the Measure it's also in a row to row context via Filter-Function, why here I cannot use naked column?

Thank you very much in advance.

Best. 

[amitchandak]

@Applicable88 , You are returning a table from filter. just like measure column can only have one value per row.

So calculation should be done assuming each row operation happens. here you are returning a table for each row

example

if( 'Calendar' [Date]>minx(relatedtable('Maintable') , 'Maintable'[BookingDate]) && 'Calendar[Day]= "Monday",1,0)

assuming Maintable is joined with calendar

[Applicable88]

@amitchandak I understand what you mean. But I'm referring to the Filter function, which you can see in the measure and also in the calculated column. In the measure 'Calendar'[Date] only works with MAX('Calendar'[Date])  whereas in the calculated column in works with a naked column like 'Calendar'[Date]>= 'Booking'[Date] 

I assume you have row by row also in the measure through Filter(), so why cannot write                'Calendar'[Date]<= 'Calendar'[Date] instead? Don't we have row context here ? Why we need to aggregate? It's a virtual table within the measure in my understanding, same as filter() in a calculated column.