I want to evaluate condition number ( http://en.wikipedia.org/wiki/Condition_number ) for general case matrix in my program with using Eigen. I have see documentation but I have not found simple way for evaluate this value.
Could you prompt the fastest way for do this?

You can use the SVD:

Code: Select all

JacobiSVD<MatrixXd> svd(A);
double cond = svd.singularValues()(0) / svd.singularValues()(svd.singularValues().size()-1);


Thank you for wery fast reply.

Yes, it is working method.
But in your way for evaluate conditon number we should evaluate all eigen numbers for AxA*.
But in case evaluate condition number we require only max and min eigen number of AxA*. So for matrices with big sizes we will do a lot of dimmy work.
Can we avoid this?