1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
|
\name{rsparsematrix}
\title{Random Sparse Matrix}
%
\keyword{array}
\keyword{distribution}
\keyword{utilities}
%
\alias{rsparsematrix}
%
\description{
Generate a random sparse matrix efficiently. The default has rounded
gaussian non-zero entries, and \code{rand.x = NULL} generates random
patter\bold{n} matrices, i.e. inheriting from \code{\linkS4class{nsparseMatrix}}.
}
\usage{
rsparsematrix(nrow, ncol, density, nnz = round(density * maxE),
symmetric = FALSE,
rand.x = function(n) signif(rnorm(n), 2), \dots)
}
\arguments{
\item{nrow, ncol}{number of rows and columns, i.e., the matrix
dimension (\code{\link{dim}}).}
\item{density}{optional number in \eqn{[0,1]}, the density is the
proportion of non-zero entries among all matrix entries. If
specified it determines the default for \code{nnz}, otherwise
\code{nnz} needs to be specified.}
\item{nnz}{number of non-zero entries, for a sparse matrix typically
considerably smaller than \code{nrow*ncol}. Must be specified if
\code{density} is not.}
\item{symmetric}{logical indicating if result should be a matrix of
class \code{\linkS4class{symmetricMatrix}}. Note that in the symmetric
case, \code{nnz} denotes the number of non zero entries of the upper
(or lower) part of the matrix, including the diagonal.}
\item{rand.x}{\code{\link{NULL}} or the random number generator for the \code{x} slot, a
\code{\link{function}} such that \code{rand.x(n)} generates a
numeric vector of length \code{n}. Typical examples are
\code{rand.x = rnorm}, or \code{rand.x = runif}; the default is nice
for didactical purposes.}
\item{\dots}{optionally further arguments passed to
\code{\link{sparseMatrix}()}, notably \code{repr}.}
}
\details{
The algorithm first samples \dQuote{encoded} \eqn{(i,j)}s without
replacement, via one dimensional indices, if not \code{symmetric}
\code{\link{sample.int}(nrow*ncol, nnz)}, then---if \code{rand.x} is
not \code{NULL}---gets \code{x <- rand.x(nnz)} and calls
\code{\link{sparseMatrix}(i=i, j=j, x=x, ..)}. When
\code{rand.x=NULL}, \code{\link{sparseMatrix}(i=i, j=j, ..)} will
return a patter\bold{n} matrix (i.e., inheriting from
\code{\linkS4class{nsparseMatrix}}).
}
\value{
a \code{\linkS4class{sparseMatrix}}, say \code{M} of dimension (nrow,
ncol), i.e., with \code{dim(M) == c(nrow, ncol)}, if \code{symmetric}
is not true, with \code{nzM <- \link{nnzero}(M)} fulfilling
\code{nzM <= nnz} and typically, \code{nzM == nnz}.
}
\author{Martin Maechler}
\examples{
\dontshow{ % for R_DEFAULT_PACKAGES=NULL
library(stats, pos = "package:base", verbose = FALSE)
}
set.seed(17)# to be reproducible
M <- rsparsematrix(8, 12, nnz = 30) # small example, not very sparse
M
M1 <- rsparsematrix(1000, 20, nnz = 123, rand.x = runif)
summary(M1)
## a random *symmetric* Matrix
(S9 <- rsparsematrix(9, 9, nnz = 10, symmetric=TRUE)) # dsCMatrix
nnzero(S9)# ~ 20: as 'nnz' only counts one "triangle"
## a random patter*n* aka boolean Matrix (no 'x' slot):
(n7 <- rsparsematrix(5, 12, nnz = 10, rand.x = NULL))
## a [T]riplet representation sparseMatrix:
T2 <- rsparsematrix(40, 12, nnz = 99, repr = "T")
head(T2)
}
|
