Fabric is Generally Available. Browse Fabric Presentations. Work towards your Fabric certification with the Cloud Skills Challenge.
in Excel, there's a handy function called CHISQ.TEST, which returns the value from the chi-squared (χ2) distribution for the statistic and the appropriate degrees of freedom.
I saw that many CHISQ-* functions are available in DAX (as shown on this Excel comparisation page), but the CHISQ.TEST function seems not to be available in DAX. We are specifically looking for this function, since we have a actual_range and expected_range in our dataset, and want to show the level of significance for differences (at <0.05).
In Excel, we do get the desired result, but I'm struggling to get the same result in DAX:
Any help would be greatly appreciated!
Solved! Go to Solution.
Since I don't know how your dimensions and data is set up, it's hard to write a formula for your situation but here's how I'd duplicate the calculation from the CHISQ.TEXT documentation you've linked.
Setup:
Gender | Description | Actual | Expected |
Men | Agree | 58 | 45.35 |
Men | Neutral | 11 | 17.56 |
Men | Disagree | 10 | 16.09 |
Women | Agree | 35 | 47.65 |
Women | Neutral | 25 | 18.44 |
Women | Disagree | 23 | 16.91 |
DAX:
ChiSqTest =
VAR ChiStat =
SUMX ( Survey, ( Survey[Actual] - Survey[Expected] ) ^ 2 / Survey[Expected] )
VAR DegreesFreedom =
( DISTINCTCOUNT ( Survey[Gender] ) - 1 ) * ( DISTINCTCOUNT ( Survey[Description] ) - 1 )
RETURN
CHISQ.DIST.RT ( ChiStat, DegreesFreedom )
Since I don't know how your dimensions and data is set up, it's hard to write a formula for your situation but here's how I'd duplicate the calculation from the CHISQ.TEXT documentation you've linked.
Setup:
Gender | Description | Actual | Expected |
Men | Agree | 58 | 45.35 |
Men | Neutral | 11 | 17.56 |
Men | Disagree | 10 | 16.09 |
Women | Agree | 35 | 47.65 |
Women | Neutral | 25 | 18.44 |
Women | Disagree | 23 | 16.91 |
DAX:
ChiSqTest =
VAR ChiStat =
SUMX ( Survey, ( Survey[Actual] - Survey[Expected] ) ^ 2 / Survey[Expected] )
VAR DegreesFreedom =
( DISTINCTCOUNT ( Survey[Gender] ) - 1 ) * ( DISTINCTCOUNT ( Survey[Description] ) - 1 )
RETURN
CHISQ.DIST.RT ( ChiStat, DegreesFreedom )
Dear AlexisOlson,
That is brilliant! It does exactly what we want! No need for a built-in CHITEST function when we can write it in DAX this way 🙂 Thank you very much 👍
Check out the November 2023 Power BI update to learn about new features.