hypervolume_prune: Removes small hypervolumes from a HypervolumeList

In dc165/Hypervolume-Dev: High Dimensional Geometry and Set Operations Using Kernel Density Estimation, Support Vector Machines, and Convex Hulls

Description

Identifies hypervolumes characterized either by a number of uniformly random points or a volume below a user-specified value and removes them from a HypervolumeList.

This function is useful for removing small features that can occur stochastically during segmentation after set operations or hole detection.

Usage

hypervolume_prune(hvlist, num.points.min = NULL, volume.min = NULL, return.ids=FALSE)

Arguments

hvlist	A HypervolumeList object.
num.points.min	The minimum number of points in each input hypervolume.
volume.min	The minimum volume in each input hypervolume
return.ids	If TRUE, returns indices of input list as well as a pruned hypervolume list

Details

Either minnp or minvol (but not both) must be specified.

Value

A HypervolumeList pruned to only those hypervolumes of sizes above the desired value. If returnids=TRUE, instead returns a list structure with first item being the HypervolumeList and the second item being the indices of the retained hypervolumes.

See Also

hypervolume_holes, hypervolume_segment

Examples

# low sample sizes to meet CRAN time requirements
data(iris)
hv1 <- hypervolume_gaussian(iris[,1:3],kde.bandwidth=0.1)
hv1_segmented <- hypervolume_segment(hv1, 
                          num.points.max=100, distance.factor=1,
                          check.memory=FALSE) # intentionally under-segment
hv1_segmented_pruned <- hypervolume_prune(hv1_segmented, 
                          num.points.min=10)
plot(hv1_segmented_pruned)

dc165/Hypervolume-Dev documentation built on Dec. 13, 2020, 6:02 p.m.