Sample data (actual dataset a few GB's in size).
Call it Table1:
| District | School | # of students |
| 1 | A | 1001 |
| 1 | B | 2163 |
| 1 | C | 1345 |
| 1 | D | 546 |
| 2 | E | 456 |
| 2 | F | 125 |
| 2 | G | 567 |
| 3 | H | 1020 |
| 3 | I | 1150 |
My final goal is to have a table with the mean and standard deviation of school sizes for the top nth percentile for each district.
So basically:
1) Take the top nth percentile (let's say top 80th as an example) school size for each district
2) Find the mean and variance or standard deviation school size for that top 80th percentile group for each district
Final result might look like (didn't calculate values based on sample data):
| District | Mean # of students (top 90th percentile) | Var # of students (top 90th percentile) |
| 1 | #### | #### |
| 2 | #### | #### |
| 3 | #### | #### |
So far I've only managed to use a measure (see below) to find the value of the top 90th percentile and use a Table visualization to see this measure for each district, but it's far from my goal.
i.e. Measure=PERCENTILE.INC(Table1[# of students],0.9)
