SparseM.ontology {SparseM}R Documentation

Sparse Matrix Class


This group of functions evaluates and coerces changes in class structure.


as.matrix.csr(x, nrow = 1, ncol = 1, eps = .Machine$double.eps), nrow = 1, ncol = 1, eps = .Machine$double.eps)
as.matrix.ssr(x, nrow = 1, ncol = 1, eps = .Machine$double.eps)
as.matrix.ssc(x, nrow = 1, ncol = 1, eps = .Machine$double.eps)
is.matrix.csr(x, ...), ...)
is.matrix.ssr(x, ...)
is.matrix.ssc(x, ...)


x is a matrix, or vector object, of either dense or sparse form
nrow number of rows of matrix
ncol number of columns of matrix
eps A tolerance parameter: elements of x such that abs(x) < eps set to zero. This argument is only relevant when coercing matrices from dense to sparse form. Defaults to eps = .Machine$double.eps
... other arguments


The function acts like matrix to coerce a vector object to a sparse matrix object of class matrix.csr. This aspect of the code is in the process of conversion from S3 to S4 classes. For the most part the S3 syntax prevails. An exception is the code to coerce vectors to diagonal matrix form which uses as(v,"matrix.diag.csr". The generic functions coerce a matrix x into a matrix of storage class The argument matrix x may be of conventional dense form, or of any of the four supported classes: matrix.csr,, matrix.ssr, matrix.ssc. The generic functions evaluate whether the argument is of class The function as.matrix transforms a matrix of any sparse class into conventional dense form. The primary storage class for sparse matrices is the compressed sparse row matrix.csr class. An n by m matrix A with real elements a_{ij}, stored in matrix.csr format consists of three arrays:

  • ra: a real array of nnz elements containing the non-zero elements of A, stored in row order. Thus, if i<j, all elements of row i precede elements from row j. The order of elements within the rows is immaterial.
  • ja: an integer array of nnz elements containing the column indices of the elements stored in ra.
  • ia: an integer array of n+1 elements containing pointers to the beginning of each row in the arrays ra and ja. Thus ia[i] indicates the position in the arrays ra and ja where the ith row begins. The last, (n+1)st, element of ia indicates where the n+1 row would start, if it existed.

    The compressed sparse column class is defined in an analogous way, as are the matrix.ssr, symmetric sparse row, and matrix.ssc, symmetric sparse column classes.


    as.matrix.ssr and as.matrix.ssc should ONLY be used with symmetric matrices.


    Koenker, R and Ng, P. (2002). SparseM: A Sparse Matrix Package for R,

    See Also

    SparseM.hb for handling Harwell-Boeing sparse matrices.


    n1 <- 10
    p <- 5
    a <- rnorm(n1*p)
    a[abs(a)<0.5] <- 0
    A <- matrix(a,n1,p)
    B <- t(A)%*%A
    A.csr <- as.matrix.csr(A) <-
    B.ssr <- as.matrix.ssr(B)
    B.ssc <- as.matrix.ssc(B)
    is.matrix.csr(A.csr) # -> TRUE # -> TRUE
    is.matrix.ssr(B.ssr) # -> TRUE
    is.matrix.ssc(B.ssc) # -> TRUE
    as.matrix.csr(rep(0,9),3,3) #sparse matrix of all zeros
    as(4,"matrix.diag.csr") #identity matrix of dimension 4

    [Package SparseM version 0.54 Index]