R/to_sparse.R

In RajLabMSSM/echoLD: echoverse module: LD downloading and processing

#' Convert to sparse
#' 
#' Convert a \link[base]{matrix} / \link[base]{data.frame} to a sparse matrix.
#' If \code{X} is already a sparse matrix, it simply returns \code{X} directly.
#' @param X A \link[base]{matrix} / \link[base]{data.frame}.
#' @param verbose Print messages.
#' @returns Sparse matrix.
#' 
#' @export
#' @importFrom methods is as
#' @examples 
#' mat <- to_sparse(X = mtcars)
to_sparse <- function(X,
                      verbose = TRUE) {
    messager("Converting obj to sparseMatrix.", v = verbose)
    if (!is_sparse_matrix(X)) {
        if (methods::is(X, "data.frame") |
            methods::is(X, "data.table") ) {
            X <- methods::as(as.matrix(X), "sparseMatrix")
        } else if(methods::is(X, "matrix") |
                  methods::is(X, "Matrix")){
            X <- methods::as(X, "sparseMatrix")
        } else {
            stp <- paste("X must be a data.frame or matrix",
                         "to convert to sparse matrix.")
            stop(stp)
        }
    }
    return(X)
}