kld: Kullback-Leibler divergence between discrete probability...

In rushkin/similR: Document similarity via word vectorization

Description

Kullback-Leibler divergence between discrete probability distributions KLD(p||q)

Usage

kld(p, q)

Arguments

p	first probability distribution ("to which"), a numeric vector, no NAs
q	second probability distribution ("from which"), a numeric vector of same length as p, no NAs

Details

Assumes that both p and q are normalized (sum(p)=sum(q)=1). Note that by definition the Kullback-Leibler divergence is asymmetric: KLD(p||q) is not equal to KLD(q||p).

Value

KLD(p||q), a numeric value

rushkin/similR documentation built on Sept. 26, 2019, 10:42 p.m.