DAX Coding Standards

Overview

Prioritize readability and maintainability: use VAR blocks, format with line breaks/indentation,
add comments, and prefer measures over calculated columns when possible.

Do’s & Don’ts

Do - Use VAR to compute sub-results once and RETURN a clear expression. Prefix your variables with an underscore.
- Format DAX consistently (each argument on a new line, indent filters).
- Add inline // comments for complex business rules.
- Use DIVIDE instead of / as DIVIDE allows safe handling of divide-by-zero.
- Use helper measures to avoid duplication and simplify maintenance.
- Use CALCULATE only when you need to change filter context, otherwise keep logic in base measures.
- Replace complex IF chains with SWITCH(TRUE(), …) for readability.
- Keep business logic in measures, not in calculated columns (except when needed for relationships).
- Test for blank values explicitly to avoid misleading 0s in reports.
- Document assumptions in comments (e.g., “Assumes sales amount is net of tax”).

Don’t - Don’t cram everything into one giant measure; break it down into helper/base measures.
- Don’t use FILTER(ALL(Table), …) when REMOVEFILTERS or ALLSELECTED is sufficient.
- Don’t overuse iterators (SUMX, AVERAGEX) when a simple aggregator (SUM, AVERAGE) works.
- Don’t create calculated columns for context-dependent logic (calculated columns are fixed at refresh).
- Don’t use ALL blindly; it removes filters across the board and may introduce unexpected results.

---

Practical Examples

[!Using VAR for clarity]-

[YoY Sales %] =
VAR _PrevYear =
    CALCULATE (
        [Total Sales],
        DATEADD ( 'Date'[Date], -1, YEAR )
    )
RETURN
    IF (
        _PrevYear = 0,
        BLANK(),
        DIVIDE ( [Total Sales] - _PrevYear, _PrevYear )
    )

[!Safe margin calculation]-

[Margin %] =
VAR _TotalSales = [Total Sales]
VAR _TotalCost  = [Total Cost]
RETURN
>   IF (
>       _TotalSales = 0,
>       BLANK(),
>       DIVIDE ( _TotalSales - _TotalCost, _TotalSales )
>   )

[!Switch instead of nested IF]-

[Customer Segment] =
SWITCH (
>   TRUE(),
>   [Customer Spend] > 10000, "High Value",
>   [Customer Spend] > 5000,  "Medium Value",
>   [Customer Spend] > 0,     "Low Value",
>   BLANK()
)

[!Remove filters selectively]-

[Total Sales (Ignore Product)] =
CALCULATE (
    [Total Sales],
    REMOVEFILTERS ( 'Product' )
)

[!Iterators vs aggregators]-

❌ Less efficient

[Total Sales X] =
SUMX ( 'Sales', 'Sales'[Quantity] * 'Sales'[UnitPrice] )

✅ Better (when column already exists)

[Total Sales] = SUM ( 'Sales'[SalesAmount] )

[!Explicit blank handling]-

[Return Rate %] =
VAR _Returns = [Total Returns]
VAR _Sales   = [Total Sales]
RETURN
 >  IF (
    ISBLANK ( _Sales ),
   BLANK(),
   DIVIDE ( _Returns, _Sales )
)