hypervolume_distance_point: Distance from a point to the margin of a hypervolume.

In hypervolume: High Dimensional Geometry, Set Operations, Projection, and Inference Using Kernel Density Estimation, Support Vector Machines, and Convex Hulls

Distance from a point to the margin of a hypervolume.

Description

Calculates the distance between two hypervolumes either defined as the Euclidean distance between centroids or as the minimum Euclidean distance between the random points comprising either hypervolume.

Code by Clement Violet.

Usage

hypervolume_distance_point(hv1, x, type = "minimum", 
  num.points.max = 1000, check.memory = TRUE)

Arguments

hv1	A Hypervolume object.
x	An object coercible to a matrix object.
type	If 'minimum', compute the smallest distance to the hypervolume margin; if 'maximum' calculates the greatest distance to the margin of the hypervolume.
num.points.max	The number of random points to subsample from the input hypervolume.
check.memory	If TRUE, prints expected memory usage and returns an error before allocating memory. Ignored if type='centroid'.

Details

Distance calculations scale quadratically with npmax and may be computationally costly.

Value

The distance between the two hypervolumes.

Examples

## Not run: 
data(penguins,package='palmerpenguins')
penguins_no_na = as.data.frame(na.omit(penguins))
penguins_adelie = penguins_no_na[penguins_no_na$species=="Adelie",
                                 c("bill_length_mm","bill_depth_mm","flipper_length_mm")]

hv1 = hypervolume_gaussian(penguins_adelie)

point = penguins_no_na[penguins_no_na$species=="Chinstrap",
                       c("bill_length_mm","bill_depth_mm","flipper_length_mm")][42, ]

# note that minimum distance is smaller than centroid distance as expected
hypervolume_distance_point(hv1, point, type='minimum', num.points.max=500, check.memory=FALSE)
hypervolume_distance_point(hv1, point, type='maximum', num.points.max=500, check.memory=FALSE)

## End(Not run)

hypervolume documentation built on April 4, 2025, 12:46 a.m.