How would you scale data to keep in in the range of 0 to 1, but in a non-linear way so that it skew more on the high side.
SO I can do it linearly:
x = (n - Rmin) /. (Rmax-Rmin)
So if the number is 50, and the range of the number is 25 to 75, and I want to normalize that number to between 0 to 1… The formula above would be: (50-25) / (75-25) = .5… This formula is similar to some of the fixed income calcuations…
Biut how do I skew this so that a number like 60 would result in something like .75 and a number like 70 would scale to .90… while a number like 10 would result in something like .11?
I tried doing a log (x) / 2 but it is too dramatic of a scale.