The Durand-Kerner method converges linearly when roots are repeating in an
equation. This reduces the accuracy of results by half the significant digits,
which is really really terrible. There's a modification to the Durand-Kerner
method that tracks multiple roots and can achieve a quadratic-like convergence
with them.
Check out PIERRE FRAIGNIAUD's paper at
http://www.springerlink.com/content/t72g1635574u10q3/
Original issue reported on code.google.com by rosen.di...@gmail.com on 27 Dec 2011 at 7:48
Original issue reported on code.google.com by
rosen.di...@gmail.com
on 27 Dec 2011 at 7:48