- Take a data table with two key columns and a third value.
- Let’s assume that no two pair of keys are the same.
- Make a table component where each cell is the aforementioned third value for one specific pair of keys.
- Rows and columns can be sorted same as any other collection.
The ability to automate some sort of rollup (say, a mean) in each end of the table sounds too good to be true, but I just put it out there.