[Maxima-bugs] [ maxima-Bugs-2932096 ] allroots fails for real polynomial

[SourceForge.net <no...@so...>]

Bugs item #2932096, was opened at 2010-01-14 10:10
Message generated for change (Comment added) made by rtoy
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=104933&aid=2932096&group_id=4933

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Lisp Core - Floating point
Group: None
>Status: Closed
>Resolution: Fixed
Priority: 5
Private: No
Submitted By: Raymond Toy (rtoy)
Assigned to: Nobody/Anonymous (nobody)
Summary: allroots fails for real polynomial

Initial Comment:
The real algorithm used in allroots fails:

allroots(x^6-24576*x^5+402653184*x^4-4947802415966*x^3+40532397764222976*x^2+9157742690304*x+2069067169);

No roots are found.  If we force allroots to use the complex algorithm, the roots are found.  If we multiply the polynomial by %i, causing allroots to use the complex algorithm, the roots are found once again.

But if we use the Fortran code from which allroots was derived, the 6 roots are easily found.

There is some bug in the translation.

----------------------------------------------------------------------

>Comment By: Raymond Toy (rtoy)
Date: 2010-01-16 10:38

Message:
Fixed in cpoly.lisp, rev 1.25.  The problem was the computation of the
error bound.   This has been replaced by the method used by complex
algorithm.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=104933&aid=2932096&group_id=4933