rectUnique: Determine a subset of rectangles with unique, non-overlapping...

In xcms: LC-MS and GC-MS Data Analysis

Description

Given a matrix of rectangular areas, this function determines a subset of those rectangles that do not overlap. Rectangles are preserved on a first come, first served basis, with user control over the order in which the rectangles are processed.

Usage

rectUnique(m, order = seq(length = nrow(m)), xdiff = 0, ydiff = 0)

Arguments

m	four column matrix defining rectangular areas
order	order in which matrix columns should be scanned
xdiff	maximum space between overlapping rectangles in x dimension
ydiff	maximum space between overlapping rectangles in y dimension

Details

The m matrix must contain four colums defining the position of rectangle sides in the folloing order: left, right, bottom, top. This function is currently implemented in C using a an algorithm with quadratic running time.

Value

A logical vector indicating which rows should be kept.

Author(s)

Colin A. Smith, [e-mail]

Examples

m <- rbind(c(0,4,0,3), c(1,3,2,6), c(3,6,4,6))
plot(0, 0, type = "n", xlim=range(m[,1:2]), ylim=range(m[,3:4]))
rect(m[,1], m[,3], m[,2], m[,4])
xcms:::rectUnique(m)
# Changing order of processing
xcms:::rectUnique(m, c(2,1,3))
# Requiring border spacing
xcms:::rectUnique(m, ydiff = 1)
# Allowing adjacent boxes
xcms:::rectUnique(m, c(2,1,3), xdiff = -0.00001)
# Allowing interpenetration
xcms:::rectUnique(m, xdiff = -1.00001, ydiff = -1.00001)

xcms documentation built on Nov. 8, 2020, 5:13 p.m.