Routines for accurate floating point summation.
There are a number of functions for accurate summation of float64s, including dot product functions.
These are mostly implementations of algorithms developed by Siegfried M. Rump and colleagues. Also included are some classic alogithms mentioned in the cited papers.