Package overview

sparseHessianFD-package

Estimate sparse Hessians using finite differences of gradients.

The sparseHessianFD class

The ‘workhorse’ for defining and computing sparse Hessian matrices.

sparseHessianFD

sparseHessianFD

Conversion helpers

For decomposing a sparse matrix into points or coordinates

Matrix.to.Pointers()

Extract row and column indices, pointers and values from a sparse matrix.

Matrix.to.Coord()

Row and column indices from sparse matrix.

Coord.to.Pointers()

Convert a matrix defined by row and column indices to one defined by a row- or column-oriented compression scheme.

Coloring algorithm details

For internal use. Do not call these directly

subst()

Estimate sparse Hessian

get_colors()

Vertex coloring of a sparse undirected graph

coloring()

Triangular partitioning of variables

Data and functions for examples

Functions and data used in the vignettes and examples

binary-data binary_small binary_large binary_super binary_small-data binary_large-data binary_super-data

Sample simulated data for binary choice example in vignette

binary.f() binary.grad() binary.hess()

Binary choice example

Deprecated

Sym.CSC.to.Matrix() Coord.to.Sym.Pattern.Matrix() Coord.to.Pattern.Matrix() new.sparse.hessian.obj() sparseHessianFD.new()

Deprecated functions