rd {divo}

rd Renyi's Divergence

Description

The Renyi divergence (RD) is a measure of similarity between two discrete probability distributions. The Renyi divergence is non-negative, not symmetric, and is not defined when there is no common support between two distributions RD is parameterized by a single non-negative parameter which may be used to adjust the relative contributions of small and large probabilities to its overall value. RD is a generalization of the Kullback-Leibler divergence. For details, see Rempala and Seweryn (2013).

Usage

rd(x, alpha = 0.5, CI = 0.95, resample = 100, graph = FALSE, csv_output = FALSE, 
PlugIn = FALSE, size = 1, CVG = FALSE, saveBootstrap = FALSE)

Arguments

x

a matrix containing input populations

alpha

Renyi's Divergence index of order alpha < 1 puts more weight on the rare species and the I Index of order alpha > 1 puts more weight on the abundant ones, default = 1

CVG

Renyi's Divergence index of order alpha = coverage. If CVG = TRUE argument alpha is ignored; default = FALSE

CI

Confidence Interval default = 0.95, range (0, 1)

resample

number of repetitions, default = 100

graph

default = FALSE, plots the results of hierarchical clustering of pairwise analysis of Renyi's Divergence; graph = 'fileName' user-defined output file name

csv_output

save the result of the analysis as .CSV file, default = FALSE; csv_output = 'fileName' user-defined output file name

PlugIn

standard plug-in estimator, default = FALSE

size

resampled fraction of the population, default = 1 (actual size of populations). The value should not be smaller than 10% of population (size = 0.1)

saveBootstrap

Saves bootstrap result to a file. Use saveBootstrap = TRUE to save bootstrap results to a Bootstrap folder in current directory; saveBootstrap = 'FolderName' - saves bootstrap results to user-named folder

Author(s)

Christoph Sadee, Maciej Pietrzak, Michal Seweryn, Cankun Wang, Grzegorz Rempala
Maintainer: Maciej Pietrzak [e-mail]

References

Rempala G.A., Seweryn M. (2013) Methods for diversity and overlap analysis in T-cell receptor populations. J Math Biol 67:1339-68

Examples

data(TCR.Data)
result <- rd(x, resample = 25, alpha=0.5)