The calculated dimension for CaseID gives us the claims corresponding to the median values in each quartile. As you can see, a Match() function is being used to match the CaseID with each of the four expressions within. Let's decipher the first expression inside the Match() function:

'$(=FirstSortedValue(distinct 
{<Value={"<=$(=Median({<Value={'>=$(=fractile(Value, 
   0))<=$(=fractile(Value, 0.25))'}>} Value))"}>} CaseID, -Value))' 

The details of the expressions are as follows:

• The innermost set gives us the range of values that are between the 0th quartile value and the twenty-fifth quartile value
• The Median() function then gives us the value that lies at the median of this range
• The FirstSortedvalue() returns the value of the output field (CaseID) based on the sorted values of the value field

In situations where the number of claims in any quartile is an even number, there will be two claims that will correspond to the median values. In such a scenario, we want to select only the claim that is higher in the sorting order. Hence, we use a -Value as the sort weight.