Float representation

The float is represented as follows :

sign bint , bint bint ...

bint represents a n bits bitset table.

Therefore the max value ranges is +/- (2^32)-(2^-(n-1)*32) with a variable precision defined by the bint table size.