Clustering with Power BI

[Anonymous]

Hello

I have data fields like:

Gender, Height, Weight, Disease

with values like:

M/F, 110-220cm, 40-250kg, Y/N

How can I identify distinct groups of the above and their probability to have Disease=Y/N, i.e. perform clustering?

I am looking to identify any possible combinations e.g. Weight=120-150kg and Gender=F and Height=150-160cm, can have 80% probability for Disease=Y etc, in addition to single variables i.e. Gender=F alone having 20% probability of Disease=Y.

Also, being able to choose multiple outputs would also be good and of course various types of variables/outputs i.e. numerical, binary, independent or not etc.

Thanks!

[Anonymous]

Thanks all but I am talking about the statistical machine learning modelling called clustering rather than simply grouping.

[v-luwang-msft]

Hi @Anonymous ，

You could create measure like the below :

test = CALCULATE(COUNTROWS('Table'),ALLSELECTED('Table'),'Table'[Disease]="Y")

Then create visual :

You can choose the range you want to get the results you want by adjusting according to the slicer.

Did I answer your question? Mark my post as a solution!

Best Regards

Lucien

[KerKol]

Ben's blog has a piece on clustering:

Clustering in Power BI using R - Ben's Blog (datakuity.com)

[amitchandak]

@Anonymous , if all these are columns in same table, you can switch true() and create a new column

Swich( True() ,

[Weight]="120-150kg"  && [Gender]="F" && [Height]="150-160cm" , "Y"

//Or you can >= and <= [Height]>=150 && [Height] <=160

// Add others  ,

"N") //else value

[amitchandak]

@Anonymous , if all these are columns in same table, you can switch true() and create a new column

Swich( True() ,

[Weight]="120-150kg"  && [Gender]="F" && [Height]="150-160cm" , "Y"

//Or you can >= and <= [Height]>=150 && [Height] <=160

// Add others  ,

"N") //else value