How is a table value handled/processed in the expression of CALCULATE?

[Anonymous]

I'm hoping someone can help me understand the logic of how the expression in a CALCULATE function works when there's only inserted a table column in it. Beneath I've created an example:

Fact table:

Type:        Value:
A                  1

B                  5

C                  2

Slicer:
A slicer that filters out "type = C".

Measures:

Following measures is the source of my question:

measure1 = 
   var ValuesColumn = VALUES('Table'[Value])
   Return
   CALCULATE(COUNTA('Table'[Value];ValuesColumn)

measure2 = 
var ValuesColumn = VALUES('Table'[Value])
Return
CALCULATE(COUNTA('Table'[Value];FILTER('Table';'Table'[Value] in ValuesColumn)

I know this is a very simplified setup but my question is basically how these 2 approaches are handled/processed? And do they even yield a different result?

[v-shex-msft]

Hi @Anonymous ,

I'd like to suggest you take a look at following blog about variable use in Dax calculation:

Variables in DAX 
Regards,

Xiaoxin Sheng

[d_gosbell · Accepted solution]

In your example with just a single table both measures will returnt the same value, but the filter context of both is slightly different so in a more complicated scenario the results could be different.

Basically in both measures the variable ValuesColumn will return a single column table with a single value of "2"

measure1 says to count the values in the Table[Value] column where  the value in Table[Value] is "2" (from the variable)

measure2 also says to count the values in the Table[Value column, but the filter this time is on the entire "Table" table (so both the Type and Value columns) so it is effectively filtered for Type = "C" and Value = 2.

In this single table example both measure1 and measure2 will return the same value, but the extra columns in the filter condition in measure2 could change things in more complicated scenarios. 

[Anonymous]

Thank you for the explanation and examples @d_gosbell