Carl Banks wrote: > If you don't have a great need for speed, you can accomplish this > easily with the linear algebra module of Numeric/numarray. Suppose > your quintic polynomial's in the form > > a + b*x + c*x**2 + d*x**3 + e*x**4 + x**5 > > The roots of it are equal to the eigenvalues of the companion matrix: > > 0 1 0 0 0 > 0 0 1 0 0 > 0 0 0 1 0 > 0 0 0 0 1 > -a -b -c -d -e
The method "zeros()" in Scientific.Functions.Polynomial uses exactly that trick for finding the zeros of a general polynomial. If you need to do more with polynomials than just finding the zeros, the Polynomial class is probably better than an on-the-spot solution. Root finding through eigenvalues is not the fastest method, but it's simple and stable, and not terribly bad either. Sorry for not making that comment earlier, I don't have the time to follow this list at the moment (to my great regret), but I was made aware of this thread through PythonURL. Konrad. -- http://mail.python.org/mailman/listinfo/python-list