Value = Expm ( Number AS Float ) AS Float
Computes Expm(x) = Exp(x) - 1.
This function is supposed to keep the precision when Number is near zero.