accurate 0.1.1

(more or less) accurate floating point algorithms
Documentation