Virtual Class of Symmetric Matrices in Package Matrix

The virtual class of symmetric matrices, "symmetricMatrix", from the package Matrix contains numeric and logical, dense and sparse matrices, e.g., see the examples with the “actual” subclasses.

The main use is in methods (and C functions) that can deal with all symmetric matrices, and in as(*, "symmetricMatrix").

Slots

uplo:

Object of class "character". Must be either "U", for upper triangular, and "L", for lower triangular.

%% below {Dim, Dimnames} work around Slot parsing buglet (< 2.2.0) %% \item{\code{Dim},\code{Dimnames}:}{The dimension (a length-2

Dim, Dimnames:

The dimension (a length-2 "integer") and corresponding names (or NULL), inherited from the Matrix, see there. See below, about storing only one of the two Dimnames components.

factors:

a list of matrix factorizations, also from the Matrix class.

Extends

Class "Matrix", directly.

Methods

dimnames

signature(object = "symmetricMatrix"): returns symmetric dimnames, even when the Dimnames slot only has row or column names. This allows to save storage for large (typically sparse) symmetric matrices.

isSymmetric

signature(object = "symmetricMatrix"): returns TRUE trivially.

There's a C function symmetricMatrix_validate() called by the internal validity checking functions, and also from getValidity(getClass("symmetricMatrix")).

Validity and dimnames

The validity checks do not require a symmetric Dimnames slot, so it can be list(NULL, <character>), e.g., for efficiency. However, dimnames() and other functions and methods should behave as if the dimnames were symmetric, i.e., with both list components identical.

See also

isSymmetric which has efficient methods (isSymmetric-methods) for the Matrix classes. Classes triangularMatrix, and, e.g., dsyMatrix for numeric dense matrices, or lsCMatrix for a logical sparse matrix class.

Examples

## An example about the symmetric Dimnames:
sy <- sparseMatrix(i= c(2,4,3:5), j= c(4,7:5,5), x = 1:5, dims = c(7,7),
                   symmetric=TRUE, dimnames = list(NULL, letters[1:7]))
sy # shows symmetrical dimnames
#> 7 x 7 sparse Matrix of class "dsCMatrix"
#>   a b c d e f g
#> a . . . . . . .
#> b . . . 1 . . .
#> c . . . . . 3 .
#> d . 1 . . 4 . 2
#> e . . . 4 5 . .
#> f . . 3 . . . .
#> g . . . 2 . . .
sy@Dimnames  # internally only one part is stored
#> [[1]]
#> NULL
#> 
#> [[2]]
#> [1] "a" "b" "c" "d" "e" "f" "g"
#> 
dimnames(sy) # both parts - as sy *is* symmetrical
#> [[1]]
#> [1] "a" "b" "c" "d" "e" "f" "g"
#> 
#> [[2]]
#> [1] "a" "b" "c" "d" "e" "f" "g"
#> 
showClass("symmetricMatrix")
#> Virtual Class "symmetricMatrix" [package "Matrix"]
#> 
#> Slots:
#>                                               
#> Name:       uplo   factors       Dim  Dimnames
#> Class: character      list   integer      list
#> 
#> Extends: 
#> Class "compMatrix", directly
#> Class "Matrix", by class "compMatrix", distance 2
#> Class "replValueSp", by class "Matrix", distance 3
#> 
#> Known Subclasses: 
#> Class "nsyMatrix", directly
#> Class "nspMatrix", directly
#> Class "lsyMatrix", directly
#> Class "lspMatrix", directly
#> Class "dsyMatrix", directly
#> Class "dspMatrix", directly
#> Class "nsCMatrix", directly
#> Class "nsRMatrix", directly
#> Class "nsTMatrix", directly
#> Class "lsCMatrix", directly
#> Class "lsRMatrix", directly
#> Class "lsTMatrix", directly
#> Class "dsCMatrix", directly
#> Class "dsRMatrix", directly
#> Class "dsTMatrix", directly
#> Class "dpoMatrix", by class "dsyMatrix", distance 2
#> Class "dppMatrix", by class "dspMatrix", distance 2
#> Class "corMatrix", by class "dsyMatrix", distance 3
#> Class "pcorMatrix", by class "dspMatrix", distance 3

## The names of direct subclasses:
scl <- getClass("symmetricMatrix")@subclasses
directly <- sapply(lapply(scl, slot, "by"), length) == 0
names(scl)[directly]
#>  [1] "nsyMatrix" "nspMatrix" "lsyMatrix" "lspMatrix" "dsyMatrix" "dspMatrix"
#>  [7] "nsCMatrix" "nsRMatrix" "nsTMatrix" "lsCMatrix" "lsRMatrix" "lsTMatrix"
#> [13] "dsCMatrix" "dsRMatrix" "dsTMatrix"

## Methods -- applicaple to all subclasses above:
showMethods(classes = "symmetricMatrix")
#> Function: Cholesky (package Matrix)
#> A="symmetricMatrix"
#> 
#> Function: Schur (package Matrix)
#> x="symmetricMatrix"
#> 
#> Function: chol (package base)
#> x="symmetricMatrix"
#> 
#> Function: chol2inv (package base)
#> x="symmetricMatrix"
#> 
#> Function: coerce (package methods)
#> from="Matrix", to="symmetricMatrix"
#> from="dgCMatrix", to="symmetricMatrix"
#>     (inherited from: from="Matrix", to="symmetricMatrix")
#> from="dgTMatrix", to="symmetricMatrix"
#>     (inherited from: from="Matrix", to="symmetricMatrix")
#> from="diagonalMatrix", to="symmetricMatrix"
#> from="dpoMatrix", to="symmetricMatrix"
#> from="dsCMatrix", to="symmetricMatrix"
#> from="dsRMatrix", to="symmetricMatrix"
#> from="dsTMatrix", to="symmetricMatrix"
#> from="dsyMatrix", to="symmetricMatrix"
#> from="lgeMatrix", to="symmetricMatrix"
#>     (inherited from: from="Matrix", to="symmetricMatrix")
#> from="lsCMatrix", to="symmetricMatrix"
#> from="lsTMatrix", to="symmetricMatrix"
#> from="lspMatrix", to="symmetricMatrix"
#> from="lsyMatrix", to="symmetricMatrix"
#> from="matrix", to="symmetricMatrix"
#> from="ngeMatrix", to="symmetricMatrix"
#>     (inherited from: from="Matrix", to="symmetricMatrix")
#> from="nsCMatrix", to="symmetricMatrix"
#> from="nsTMatrix", to="symmetricMatrix"
#> 
#> Function: dimnames (package base)
#> x="symmetricMatrix"
#>