revisit EFloat implementation #72

mmp opened this Issue May 30, 2016 · 0 comments


None yet

1 participant

mmp commented May 30, 2016

The current implementation is based on maintaining an interval, which is fine, but has diverged (already) from the book text.

Krämer's paper "A priori worst-case error bounds for floating-point computation" ( outlines how to compute bounds using a computed value and an error interval; it would be nice to use his method to fix the bugs that caused the original shift to the interval-based approach (46de256) and get things back in better sync with the book text.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment