A function to do incomplete LU decomposition with a drop tolerance?
- To: mathgroup at smc.vnet.net
- Subject: [mg123287] A function to do incomplete LU decomposition with a drop tolerance?
- From: "Nasser M. Abbasi" <nma at 12000.org>
- Date: Wed, 30 Nov 2011 03:21:55 -0500 (EST)
- Delivered-to: l-mathgroup@mail-archive0.wolfram.com
- Reply-to: nma at 12000.org
Version 8.0.4 I need to use ILU as a preconditioner for a conjugate gradient solver I am writing in Mathematica. But looking around, I am not able to find one in Mathematica. Only reference I found is here http://reference.wolfram.com/mathematica/tutorial/SomeNotesOnInternalImplementation.html Where it say in the above the following: "For sparse arrays, LinearSolve uses UMFPACK multifrontal direct solver methods and with Method uses Krylov iterative methods preconditioned by an incomplete LU factorization" But how does a user access this ILU from Mathematica? I do have the A matrix (sparse). Also, I see still that the Incomplete Cholesky decomposition does not have a drop tolerance parameter. I hope in version 9 it will. thanks, --Nasser