Vectorizing a matrix and its inverse

tda_mcvec stacks the columns of A into a single vector – TDA's mcvec, the classic matrix-algebra vec(A) operator, matching base R's as.vector(A) (R matrices are column-major already). tda_mrvec stacks the rows instead – TDA's mrvec, matching as.vector(t(A)). tda_mivec is the inverse of tda_mcvec: given a vector and a row count n, it rebuilds an n-row matrix by filling column by column – TDA's mivec, matching base R's matrix(x, nrow = n).

Usage

tda_mcvec(A, dir = tempfile("tda"))

tda_mrvec(A, dir = tempfile("tda"))

tda_mivec(x, n, dir = tempfile("tda"))

Arguments

A

a numeric matrix.

dir

working directory.

x

a numeric vector, whose length must divide evenly by n.

n

the number of rows for tda_mivec's result.

Value

tda_mcvec/tda_mrvec return a plain numeric vector; tda_mivec returns a matrix.

See also

Other matrix reshaping: tda_mag(), tda_mcath(), tda_mrsum(), tda_msort(), tda_msrow(), tda_mtrim()

Examples

A <- matrix(1:6, 2, 3)
tda_mcvec(A)               # == as.vector(A)
#> [1] 1 2 3 4 5 6
tda_mrvec(A)               # == as.vector(t(A))
#> [1] 1 3 5 2 4 6
tda_mivec(1:6, 2)          # == matrix(1:6, nrow = 2)
#>      [,1] [,2] [,3]
#> [1,]    1    3    5
#> [2,]    2    4    6