det.cut: Identify Tree Leafs Intersected by Condition(s)

In detpack: Density Estimation and Random Number Generation with Distribution Element Trees

Description

Identify distribution element tree (DET) leafs that are cut by conditions. The latter are defined in terms of positions xc along probability-space components with indices dc.

Usage

det.cut(det, xc, dc)

Arguments

det	distribution element tree object resulting from det.construct.
xc	vector with conditioning values of probability-space components listed in dc.
dc	integer vector with indices of conditioning components corresponding to xc.

Value

A vector containing the leaf indices that are cut by conditions xc of components dc is returned. If no leafs are found, the return vector has length 0.

Examples

# DET based on Gaussian data
require(stats); require(graphics)
n <- 8e4; x <- rnorm(n)
x <- matrix(c(x, x+rnorm(n,0,0.2)), ncol = 2)
det <- det.construct(t(x), lb = 0, ub = 0) # no pre-whitening
plot(x, type = "p", pch = ".", asp = 1)
# leaf elements that are cut by x1 = 2
leafs <- det.cut(det, xc = 2, dc = 1) # condition x1 = 2
# draw probability space (black) with cut leaf elements (red)
rect(det$lb[1], det$lb[2], det$ub[1], det$ub[2], border = "black")
for (k in 1:length(leafs)) {
   p <- det.de(det, leafs[k])$lb; w <- det.de(det, leafs[k])$size
   rect(p[1],p[2],p[1]+w[1],p[2]+w[2], border = "red")
}
# leafs cut by two conditions x1 = -3, x2 = -2 (blue)
leafs <- det.cut(det, xc = c(-2,-3), dc = c(2,1))
p <- det.de(det, leafs[1])$lb; w <- det.de(det, leafs[1])$size
rect(p[1],p[2],p[1]+w[1],p[2]+w[2], border = "blue")

detpack documentation built on July 24, 2019, 5:03 p.m.