accurate 0.3.0

(more or less) accurate floating point algorithms
Documentation