You can use Lerp to interpolate a number within your range. If the scale is not linear than you would need to code your own function (using a look up table, for example).
For this particular example, why not just do [float * (0.9 + float)]
EDIT: Ah, I see.
Well, in addition to mapping a range, you could also define a float curve. For a value, you specify a corresponding value… Then you just Get the value from the curve at that point (for multiplication… Or you could just simplify it and map an X value to the desired final Y value)