Back to Sparse Direct Solver (SuperLU)
slu.permutation.computeColumnPermutation
Compute a column permutation for fill-in reduction.
Syntax
slu.permutation.computeColumnPermutation(A, strategy?)
Description
Compute a column permutation for fill-in reduction. Column permutations reorder the columns of the matrix to reduce fill-in during LU factorization. Different strategies trade off between computation time and fill-in reduction quality: - **natural**: No reordering (identity permutation) - **mmd_ata**: Minimum degree on A^T * A - **mmd_at_plus_a**: Minimum degree on A^T + A - **colamd**: Column approximate minimum degree (recommended for most cases) - **metis**: METIS nested dissection (best for very large matrices)
Parameters
| Name | Description |
|---|---|
| A | - Sparse matrix in CSC format |
| strategy(optional) | - Permutation strategy |
Returns
ColumnPermutationResult