[SciPy-User] Eigen not working

[Pauli Virtanen]

Thu, 09 Jun 2011 19:12:02 +0530, Vivek Lohani wrote:
> I was trying to implement the code for diagonalizing sparse matrices
> through scipy.sparse.linalg.eigen but it turns out that for N=9 (which
> leads to
>  2^N X 2^N matrix) and the subsequent input paramter= f, which sets
>  J=-1,
> I am not getting the correct output for -0.71<g<0.71. It also happens
> for lower values of N(>=5) on my computer in a region -0.6<g<0.6. I am
> unable to understand what is going wrong because i have a numpy routine
> too which does the job correctly.

Try upgrading to Scipy 0.9 (recommended), or specify a larger `maxiter`.

In Scipy 0.8 and earlier, the Arpack eigenvalue routines essentially
left convergence checking to the user, so if `maxiter` (default 20*n)
is too small, they return non-converged results.

-- 
Pauli Virtanen