LinearConv2: Linear Convolution of a Numeric Matrix

In ocbe-uio/CellMigRation: Track Cells, Analyze Cell Trajectories and Compute Migration Statistics

Description

Performs a linear convolution of a Numeric Matrix, using a user-supplied linear kernel. The convolution can be executed in a column-wise fashion by setting the col.wise argument to TRUE. Alternatively, the convolution is performed in a row-wise fashion.

Usage

LinearConv2(x, krnl, col.wise = TRUE)

Arguments

x	numeric matrix that will be used as input for the convolution; this matrix typically corresponds to an image where signal (high values) indicates the presence of a cell or a cell-like particle
krnl	numeric vector corresponding to the kernel that will be used for the convolution. Briefly, the kernel includes the weights that will be used to compute a weighted sum at each position of the input numeric matrix
col.wise	logical; shall the linear convolution be performed in a column-wise or row-wise fashion?

Value

Linearly convoluted numeric matrix. The resulting matrix has the same dimensions of the inut matrix

Author(s)

Damiano Fantini, [e-mail]

References

https://www.data-pulse.com/dev_site/cellmigration/

Examples

graphics::par(mfrow = c(1, 2))
tmp <- vapply(
    seq_len(12),
    function(i) {(6 + abs(i - 6)) * c(seq(1,10,by=1), seq(10,1,by=-1))},
    FUN.VALUE = numeric(20)
)
cnv.tmp <- cellmigRation:::LinearConv2(tmp, c(-3, 0, 3))
graphics::image(tmp); graphics::image(cnv.tmp)

ocbe-uio/CellMigRation documentation built on Dec. 26, 2021, 10:19 p.m.