0

What is the maximum size of a sparse matrix that can be diagonalized using Numpy Linalg package?

I have been working with exact diagonalisation methods such as Lanczos and Arnoldi in the context of diagonalizing my Hamiltonian matrices.

Recently I have started to use Numpy's linalg to do these numerical calculations.

I have been wondering about, what would be the maximum size of a very sparse matrix that can be diagonalised using Numpy. Are there any hacks to get beyond a certain limit and is it possible to get the code to work faster are there any best practices on this kind of coding with numpy?

Egon Allison
  • 1,329
  • 1
  • 13
  • 22
Fracton
  • 171
  • 5
  • It's not clear whether you have the scipy sparse matrices in mind or not. But if you do, you should look at the `scipy.sparse.linalg` functions rather than the numpy.linalg ones. – hpaulj Mar 07 '19 at 07:26
  • Also see my comments for https://stackoverflow.com/q/55009318/901925 – hpaulj Mar 07 '19 at 07:29
  • Any idea what would be the maximum size of a matrix that I could diagonalise using numpy – Fracton Mar 08 '19 at 05:24
  • NumPy's diagonalization routines (and its array representation) don't take advantage of sparsity at all. – user2357112 Mar 23 '20 at 00:39

0 Answers0