Forum Discussion

SteveIOW's avatar
SteveIOW
Icon for Helper II rankHelper II
3 years ago
Solved

Using Variables in Summarize

Hi, wondering if anyone understands what is happening here.   I have built a table using summarize. To start with I used 4 variables: VAR vC2021 = CALCULATE(DISTINCTCOUNT(MainFacts[EpisodeNumber...
Need Help
  • johnt75's avatar
    johnt75
    3 years ago

    Variables in DAX aren't really variables, they're constants. They are only calculated once, when they are defined, which is why you're seeing the same wrong values for each row.

    Its also worth pointing out that it is not best practice to use SUMMARIZE to produce calculated columns as performance is poor and you can get unexpected results. It is better to use SUMMARIZE to do the grouping and then use ADDCOLUMNS to produce the calculations, so your code would become

    My Table =
ADDCOLUMNS (
    SUMMARIZE ( MainFacts, MainFacts[Provider_Code] ),
    "Count2021",
        CALCULATE (
            DISTINCTCOUNT ( MainFacts[EpisodeNumber] ),
            MainFacts[Period] = "FY2020-21"
        ),
    "NullFlag2021",
        IF (
            CALCULATE (
                DISTINCTCOUNT ( MainFacts[EpisodeNumber] ),
                MainFacts[Period] = "FY2020-21"
            ) = 0,
            1,
            0
        ),
    "Count2122",
        CALCULATE (
            DISTINCTCOUNT ( MainFacts[EpisodeNumber] ),
            MainFacts[Period] = "FY2021-22"
        ),
    "NullFlag2122",
        IF (
            CALCULATE (
                DISTINCTCOUNT ( MainFacts[EpisodeNumber] ),
                MainFacts[Period] = "FY2021-22"
            ) = 0,
            1,
            0
        )
)
johnt75's avatar
johnt75
Icon for Super User rankSuper User
3 years ago

Variables in DAX aren't really variables, they're constants. They are only calculated once, when they are defined, which is why you're seeing the same wrong values for each row.

Its also worth pointing out that it is not best practice to use SUMMARIZE to produce calculated columns as performance is poor and you can get unexpected results. It is better to use SUMMARIZE to do the grouping and then use ADDCOLUMNS to produce the calculations, so your code would become

My Table =
ADDCOLUMNS (
    SUMMARIZE ( MainFacts, MainFacts[Provider_Code] ),
    "Count2021",
        CALCULATE (
            DISTINCTCOUNT ( MainFacts[EpisodeNumber] ),
            MainFacts[Period] = "FY2020-21"
        ),
    "NullFlag2021",
        IF (
            CALCULATE (
                DISTINCTCOUNT ( MainFacts[EpisodeNumber] ),
                MainFacts[Period] = "FY2020-21"
            ) = 0,
            1,
            0
        ),
    "Count2122",
        CALCULATE (
            DISTINCTCOUNT ( MainFacts[EpisodeNumber] ),
            MainFacts[Period] = "FY2021-22"
        ),
    "NullFlag2122",
        IF (
            CALCULATE (
                DISTINCTCOUNT ( MainFacts[EpisodeNumber] ),
                MainFacts[Period] = "FY2021-22"
            ) = 0,
            1,
            0
        )
)
Hijaika's avatar
Hijaika
Frequent Visitor
3 years ago

But what if i wanted to use that variable again in a later calculation? In this case Count2122 for example.

  • johnt75's avatar
    johnt75
    Icon for Super User rankSuper User
    3 years ago

    Instead of using ADDCOLUMNS you could use a combination of GENERATE and ROW, the general pattern is something like

    My Table =
GENERATE (
    SUMMARIZE ( 'Table', 'Table'[Column 1], 'Table'[Column 2] ),
    VAR Var1 =
        CALCULATE ( SUM ( 'Table2'[Value] ) )
    VAR Var2 =
        IF (
            Var1 > 0,
            CALCULATE ( SUM ( 'Table2'[Value2] ) ),
            CALCULATE ( SUM ( 'Table2'[Value3] ) )
        )
    RETURN
        ROW ( "Var1", Var1, "Var2", Var2 )
)
    • SteveIOW's avatar
      SteveIOW
      Icon for Helper II rankHelper II
      3 years ago

      all of this is incredibly helpful. Many many thanks.