Gets the rational decimal that is closest to the given decimal value, and returns the fractional equivalent. Currently only accepts numbers between 0 and 1 (exclusive). Known float values are derived from fractions with denominators up to 2,500.