Back to Sintra/Data & Analytics
intermediate·Data & Analytics·20 min

Auto-Generate DAX Measures from Plain English

A production-ready DAX measure in VAR/RETURN style, plain-English explanations of every function, edge-case warnings, and a simplified version.

⬡
Recommended modelClaude Sonnet 4.6

Handles complex queries across large datasets

What you need to fill in

[describe the calculation in plain English][fact table name and key columns][date table name and whether it is marked as a date table][related dimension tables][existing base measures to reference]

Tools used

Claude SonnetPower BI DesktopPower BI Copilot

The prompt

"I need a DAX measure for Power BI. Here is what I want it to calculate in plain English: [Describe the calculation, e.g. 'Year-to-date revenue compared to the same period last year, expressed as a % change'] Data model context: - Fact table: [name + key columns] - Date table: [name + date column, is it marked as a date table? yes/no] - Related dimensions: [e.g. Customer, Product, Region tables] - Existing base measures I can reference: [list, e.g. [Total Revenue], [Units Sold]] Please: 1. Write the DAX measure using VAR/RETURN style 2. Explain each function in one plain-English sentence 3. Flag any edge cases (blanks, missing periods, filter context issues) 4. Provide a simplified version for a less experienced colleague"
Open in Claude

Sample output

**Production measure:** ```dax Revenue YoY % = VAR _Curr = [Total Revenue] VAR _Prev = CALCULATE([Total Revenue], SAMEPERIODLASTYEAR('Date'[Date])) RETURN DIVIDE(_Curr - _Prev, _Prev, BLANK()) ``` **Function explanations:** - `CALCULATE` — re-evaluates a measure under a modified filter context - `SAMEPERIODLASTYEAR` — shifts the date filter back exactly 12 months - `DIVIDE` — safe division that returns BLANK() instead of an error when the denominator is zero **Edge cases:** - Returns BLANK() for periods with no prior-year data — visuals handle this cleanly - Requires a marked date table; if not marked, replace with `DATEADD('Date'[Date], -1, YEAR)` - Slicer selections (e.g. Region) are automatically respected via filter context **Simplified version:** ```dax Revenue YoY % Simple = DIVIDE( [Total Revenue] - CALCULATE([Total Revenue], SAMEPERIODLASTYEAR('Date'[Date])), CALCULATE([Total Revenue], SAMEPERIODLASTYEAR('Date'[Date])) ) ```

Try with these tools

claudechatgptnotebooklm
daxpower-bimeasuresai-assisted

More in Data & Analytics

FP&A KPI Dictionary Standardiser

A single KPI dictionary: definition, formula, source system, owner, and refresh cadence per metric.

intermediate

Procurement Spend Cube & Savings Finder

A cleansed spend cube (supplier × category × BU) with ranked savings opportunities.

intermediate

Master Data Golden Record & Dedup Ruleset

A concrete matching, survivorship, and golden-record ruleset for cleaning a customer or vendor master.

advanced