|
@@ -141,6 +141,12 @@ IGL_INLINE bool igl::min_quad_with_fixed_precompute(
|
|
|
// Get transpose of Aequ
|
|
|
Eigen::SparseMatrix<T> AequT = Aequ.transpose();
|
|
|
// Compute LU decomposition
|
|
|
+ // TODO: This should be using SimplicialLDLT
|
|
|
+ // Q: Does SimplicialLDLT work on "Hermitian indefinite matrices" the way
|
|
|
+ // that matlabs ldl does?
|
|
|
+ // A: Maybe not. The eigen documentation writes, "This class provides a
|
|
|
+ // LDL^T Cholesky factorizations without square root of sparse matrices
|
|
|
+ // that are selfadjoint and positive definite"
|
|
|
bool lu_success = igl::lu_lagrange(Auu,AequT,data.L,data.U);
|
|
|
if(!lu_success)
|
|
|
{
|