ACBC-package: Analysis of convex body coverage

In mlcollyer/ACHC: Analysis of Convex Body Coverage

Value

Key functions for this package:

achc	Permutation procedure to generate many random convex hulls and measure the amount of space covered by individual hulls.
aec	Permutation procedure to generate many random ellipsoids and measure the amount of space covered by individual ellipsoids.
plot.achc	Generates plots associated with achc.
plot.aec	Generates plots associated with aec

The name, "ACBC", is an acronym for, "Analysis of Convex Body Coverage." Through the various functions in this package, one can determine if the distribution of convex bodies in a data space, represented by groups to be compared, is consistent with or different than expectation from random assembly of observations. ACBC fits a uniform grid of points to the total expanse of multidimensional space covered by observations. Convex hulls or ellipsoids for groups are described, and the frequency of 0, 1, ..., g convex bodies for g groups are measured at every point. A relative frequency distribution is generated from these frequencies. Next, the procedure is repeated many times with random assigment of observations to groups. Confidence bands are generated for the 0, 1, 2, ..., g groups. By comparing the observed distribution to the confidence limits generated, one can ascertain if the data space is meaningfully partitioned in some way.

Author(s)

Michael Collyer

mlcollyer/ACHC documentation built on May 30, 2020, 10:26 p.m.