R: Convert a matrix defined by row and column indices to one...

Coord.to.Pointers {sparseHessianFD}

R Documentation

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

Description

Returns indices and pointers that define a sparse Hessian in compressed format. Inputs are the row and column indices.

Usage

Coord.to.Pointers(
  rows,
  cols,
  dims,
  triangle = TRUE,
  lower = TRUE,
  symmetric = FALSE,
  order = c("column", "row", "triplet"),
  index1 = TRUE
)

Arguments

`rows`, `cols`	row and column indices of non-zero elements
`dims`	2-element vector for number of rows and columns.
`triangle`	Is input intended to be a triangular (TRUE) or full (FALSE) matrix. See details for how this argument is interpreted for different values of `order`.
`lower`	If `triangular` is true, this argument identifies the input matrix as lower- or upper-triangular. This argument is ignored if `triangle` is FALSE.
`symmetric`	If TRUE, and matrix is triangular, then the matrix will be treated as symmetric, with only the triangular elements provided. If matrix is neither triangular nor symmetric, then symmetric=TRUE will probably trigger an error.
`order`	Determines the indexing/compression scheme for the output matrix. Use "triplet" to get row and column indices. Defaults to the same class as M.
`index1`	TRUE if using 1-based indexing. FALSE for 0-based indexing.

Details

triangle and order have the following interpretation:

triangle=TRUE: Input rows and cols represent lower or upper triangle of a matrix. If order="symmetric", then the output list will be for a full, symmetric matrix. Otherwise, the output list will be for only the lower or upper triangle. Any elements outside of the specified triangle will trigger an error.
triangle=FALSE: Input rows and cols represent a full matrix. If that matrix is not symmetric, then order=="symmetric" will trigger an error.

If symmetric=FALSE and order='triplet', the output list should contain the same row and column indices as the input list.

Value