How to use incomplete LU factorization in Eigen?

Question

I have read the docs of IncompleteLUT

And show the detail here:

• Classes

   struct keep_diag 

• Public Member Functions

   template< typename MatrixType > IncompleteLUT< Scalar > & compute (const MatrixType &amat) ComputationInfo info () const` void setDroptol (const RealScalar &droptol)` void setFillfactor (int fillfactor)` 

But I am puzzling how to use the method to get the result matrix of incomplete LU factorization. Thank you!

Answer 1

The document in development branch is a little bit more clear.

This class follows the sparse solver concept .

https://eigen.tuxfamily.org/dox-devel/classEigen_1_1IncompleteLUT.html

The class IncompleteLUT follows the concept of Sparse Solver. All Sparse Solvers have the same interface as shown in the linked tutorial. Basically you only need to replace the SolverClassName with the desired solver name , in this case IncompleteLUT .

https://eigen.tuxfamily.org/dox-devel/group__TopicSparseSystems.html#TutorialSparseSolverConcept