accurate 0.4.0

(more or less) accurate floating point algorithms
Documentation