R/make_functions.R

In smnet: Smoothing for Stream Network Data

#' @importFrom methods as
#' @importFrom spam as.spam.dgCMatrix
#' @importFrom spam as.spam
#' @importFrom spam spam

# sparse matrix version of the crossprod function
crossprodspam <- function(X){
  t(X) %*% X
}

# create an empty sparse matrix of class spam
make_sparse <- function(ncol, nrow = ncol) spam(0, nrow = nrow, ncol = ncol)

# create a square matrix of 1s of class spam (useful for block kronecker operations)
make_one <- function(ncol, nrow = ncol) spam(1, nrow = nrow, ncol = ncol)

# coerce to spam with given tolerance level  
make_spam <- function(M){
  if(class(M) == "matrix") as.spam(M)
  if(class(M) == "spam") as.spam(M)
  else as.spam(as.spam.dgCMatrix(as(M, "dgCMatrix")))
}