Avoiding Poles in the Complex Plane
- From: wesh <weshdrop@xxxxxxx>
- Date: Thu, 18 May 2006 23:24:09 GMT
I have functions defined by nonlinear differential equations. I know the
initial conditions at exactly one point, z0, of the complex plane. The
complex plane is covered with an infinite lattice of simple poles, whose
position is unknown. The goal is to generate a contour map of the
absolute value of the functions over a portion of the complex plane
which includes some of the poles.
To find the value at the point z1, I integrate along the line
z0 + (z1-z0)t,
or t going from 0 to 1. This works fine except when the line passes too
close to a pole causing the numerical integrate to fail. This eliminates
a wedge of the plane behind the pole from the plot.
I need some way to dynamically adjust the trajectory so as to reach the
point z but miss any poles that would disrupt the integration.
As a concrete example consider
c'[z] = -s[z]^6,. c[0] = 1,
s'[z] = c[z]^6, s[0] = 0.
these functions generate a lattice of poles with a seven fold symmetry.
I would like to integrate along some path
z[t] = u[t] + i v[t], z[0] = z0, z[1] = z1
such that a function similar to
|z - z1| [a + (1-a)(|c[z]| + |s[z]|)],
remains below some bound. Clearly, a set of straight lines can be found
by stopping the integration and inserting a dog leg when necessary. What
would be nice is differential formulation that would allow an
uninterrupted integration from z0 to z1.
.
- Follow-Ups:
- Re: Avoiding Poles in the Complex Plane
- From: Julian V. Noble
- Re: Avoiding Poles in the Complex Plane
- From: Alan
- Re: Avoiding Poles in the Complex Plane
- Prev by Date: Re: B splines problem
- Next by Date: piecewise linear 1D finite element method
- Previous by thread: DASPK problem with zero solution components
- Next by thread: Re: Avoiding Poles in the Complex Plane
- Index(es):
Relevant Pages
|