v0.3.1

• Added the AMD algorithm to generate an efficient permutation vector to permute matrices prior to factorisation. The functions amd_permute() in the sparse class will generate the ordering and permute() and permute_inv() will return the ordering and its inverse respectively. The SparseChol class will use the ordering if it exists in the associated sparse class. For R a function amd_order() has been added to generate the order for a given matrix.

v0.2.2

• Added support for Eigen matrices. Sparse class can be initialised using an Eigen matrix, and the function sparse_to_dense() will convert to an Eigen matrix. dense_to_sparse() will convert an Eigen matrix to sparse matrix.
• Added more operators (where dblvec = std::vector<double>:
  • %= dblvec right multiplication by a diagonal matrix represented as a vector
  • sparse * sparse mutiplication
  • sparse * dblvec for sparse matrix times a vector
  • sparse % dblvec for sparse multiplication with a diagonal matrix represented as a vector
  • sparse + sparse
  • sparse * Eigen::Dense
  • Eigen::Dense * sparse
  • sparse * VectorXd
  • sparse % VectorXd where the vector represents a diagonal matrix
• Added function identity to return a sparse identity matrix
• Added both row and column major representations to sparse class
• Added member function insert to sparse class to insert an element at the given position

v0.1.1

First release