[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference rusure::math

Title:Mathematics at DEC
Moderator:RUSURE::EDP
Created:Mon Feb 03 1986
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2083
Total number of notes:14613

1414.0. "EXACT Polynomial roots with symbolic arithmatic?" by FMCSSE::HEINTZE () Thu Apr 04 1991 23:13

    I was thinking about the maple program.  I recal that it would find
    roots of polynomials:   If you specified integer coefficients it used
    rational arithmatic to give the precise roots - absolutely no round off
    error!   They do this by never performing a division or a root
    operation - instead they store the numerator and denominator and
    proceed symbolically.
    
    I'm aware of the Jenkins Traub algorithm - I figured that was the last
    word on polynomial roots but this uses real number representation (real
    in the context of FORTRAN).
    
    What is the algorithm that MAPLE uses to find the exact roots of
    polynomials?
T.RTitleUserPersonal
Name
DateLines
1414.1In a word, NOCIVAGE::LYNNLynn Yarbrough @WNP DTN 427-5663Fri Apr 05 1991 11:1219
>    I'm aware of the Jenkins Traub algorithm - I figured that was the last
>    word on polynomial roots but this uses real number representation (real
>    in the context of FORTRAN).
    
>    What is the algorithm that MAPLE uses to find the exact roots of
>    polynomials?

There is a theoretical limit to the *exact* solution of nth-degree 
polynomials; regrettably, that limit is n=4. Beyond that, the solutions 
cannot be expressed as radicals, i.e. there *are no* exact solutions, in 
general. What MAPLE will do is break down the problem and report the
results in the form 
	RootOf(expression)
which begs the question a lot. Jenkins-Traub is a good computational 
rootfinder but makes no pretensions about finding exact roots.

The algorithms MAPLE uses for n = 2,3,4 are those found by Cardan et al a
couple of centuries ago, and are given in most high school or college
algebra texts. 
1414.2finding all roots for ill-condition polySMAUG::ABBASIFri Apr 05 1991 12:5117
    ref .0
    in Maple also, for an ill-condition problem such as
    
        4     -8
    (x-2) - 10    = 0
    
    to find roots you should use fsolve, as it tries to find all roots for
    an ill-conditioned problem, but it still could miss some. this  relates
    to the nature of the problem not the method.
    
    FUNCTION : fsolve - solve using floating-point arithmetic
    
    For  a general equation, fsolve attempts to compute a single real root.  
    However  for  polynomials it will compute all real (non-complex) roots, 
    although  exceptionally  ill-conditioned  polynomials  may  cause  fsolve 
    to miss some roots.