A question about matrix eigenvalues.

Demidov

Hello guys! Thank you all for that valuable discussion on my previous question.

At the moment I'm trying to solve eigenvalue and eigenvector problem for that great matrix that offen results from elliptic equations. Of course its impossible to solve it using exact methods (and I dont need all the eigen values). So, I would like to use another methods.

At the moment I use: 1) FOR approximate finding eigenvalues and eigenvectors of the sparse matrix - the reverse iterations method 2) FOR approximate solving linear equation system in this method - the conjugate gradients method.

What do you think of this approach? Which methods you prefer to use in case the matrix is very large and you dont need to find all the eigenvalues?