Cosine normalization of a matrix

Normalizes a numeric matrix dividing each row (if rows=TRUE) or column (if rows=FALSE) by their L2 norm. Thus, each row (or column) has unit norm.

Usage

cosnormX(X, rows = TRUE)

Arguments

X

Numeric matrix or data.frame of any size.

rows

If TRUE, the operation is done by row; otherwise, it is done by column. (Defaults: TRUE).

Value

Cosine-normalized X.

Examples

dat <- matrix(rnorm(50),ncol=5,nrow=10)
cosnormX(dat)
#>              [,1]        [,2]        [,3]        [,4]        [,5]
#>  [1,]  0.52052548  0.50768215  0.34246093 -0.33101576 -0.49443012
#>  [2,] -0.14652804  0.79381468  0.53119300  0.25422535 -0.03989044
#>  [3,]  0.21677075  0.07554259 -0.59197766 -0.71061572  0.30313614
#>  [4,]  0.56194932 -0.08815004  0.17431695 -0.75487916 -0.27606809
#>  [5,]  0.77088910  0.02774875 -0.18482886 -0.16575575  0.58593799
#>  [6,]  0.18845708 -0.15921689  0.75203665  0.61099148  0.01625403
#>  [7,]  0.61654194 -0.11129243 -0.71211104  0.28971803 -0.12826287
#>  [8,] -0.06215042 -0.17944582 -0.23995466  0.02374133 -0.95173244
#>  [9,]  0.06756910 -0.32619437 -0.08038939 -0.61302705  0.71187571
#> [10,]  0.37252294  0.16523170 -0.69904592  0.43050287 -0.39990901