Posted by Paul on April 03, 1998 at 07:53:30:

Hi,

my problem is:

I have got a set of linear equations

Cx = 0

with C is the singular matrix and x is the non-trivial

solution vector. Additionally, C is a function of one

or two scalar variables a and/or b:

C = C(a,b)

The exact values of these variables are unknown and have

to be find out by iterations. And this causes the problem:

I need an appropriate measure for how good the variables

values fit. In other words, I must be able to compare if

C(a1,b1) is stronger singular then C(a2,b2).

I have tested as a measure the smallest singular value,

but that didnt work.

Maybe anybody have got an better idea.

Thanks

Paul