Estimate Hessian using triangular subsitution algorithm
subst(Y, colors, jCol, ipntr, delta, nvars, nnz)
A sparse Hessian of class dgCMatrix.
Matrix of finite differences of gradients
Vector of length nvars that identifies color of each variable
Column indices and row pointers for non-zero elements of lower triangle of Hessian (row-oriented compressed format).
Perturbation factor used to compute finite differences of gradients.
Dimension of Hessian (number of variables)
Number of non-zero elements in the lower triangle of the Hessian.
For internal use. You should not have to call this function directly.