Hi all
I've been trialling, researching, and going round in circles for weeks on this challenge and finally admitted I need some help from this community.
I have a data set containing student data. The original data has a row per student, and multiple columns for each personal characteristic (or their enrolled course characteristics). I need to be able to show combinations of these characteristics, and show how many students there are in each combination. The solution needs to be dynamic, catering for any number of characteristic types (generally will be between 5 and 20).
So far, I've pivotted the characteristic columns into Attribute and Value, and created a fact table, and Attribute-Value lookup table.
I've also created a Person-AttributeList table, with concatenated list of all Attribute-Values per student.
I've also used clusters on the Person-AttributeList which is showing clusters for ALL Person-Attributes, however I need to be able to select just one or more Attributes, and see a breakdown of number of people per Attribute-Value combination.
I've created a sample dataset and pbix file which is linked below.
In the sample, there are just 10 people, each with Attribute-Values for Age Group, Gender, Glasses, and Hair Colour. I want to be able to select (for example) Gender and Age Group in an Attribute slicer, and see how many people are "Male, 10-19 years", "Female, 10-19 years", "Male, 20-29 years" etc. This needs to not only show in a table, but also be used as a legend for charts too (and this is one area where I have got stuck as Measures can't be used for Legend).
Many thanks for any advice, links, suggestions or solutions anyone can assist me with 🙂
Attribute Permutations PBIX
