kullback_leibler_cont_appr: Calculate a Continuous Approximation of the Kullback-Leibler...

In oHMMed: HMMs with Ordered Hidden States and Emission Densities

Calculate a Continuous Approximation of the Kullback-Leibler Divergence

Description

Calculate a Continuous Approximation of the Kullback-Leibler Divergence

Usage

kullback_leibler_cont_appr(p, q)

Arguments

p	(numeric) probabilities
q	(numeric) probabilities

Details

The continuous approximation of the Kullback-Leibler divergence is calculated as follows:

\frac{1}{n}\sum_{i=1}^n\big[\log(p_i) p_i - \log(q_i) p_i \big]

Value

Numeric vector

Examples

# Simulate n normally distributed variates
n <- 1000
dist1 <- rnorm(n)
dist2 <- rnorm(n, mean = 0, sd = 2)
dist3 <- rnorm(n, mean = 2, sd = 2)

# Estimate probability density functions
pdf1 <- density(dist1)
pdf2 <- density(dist2)
pdf3 <- density(dist3)

# Visualise PDFs
plot(pdf1, main = "PDFs", col = "red", xlim = range(dist3))
lines(pdf2, col = "blue")
lines(pdf3, col = "green")

# PDF 1 vs PDF 2
kullback_leibler_cont_appr(pdf1$y, pdf2$y)

# PDF 1 vs PDF 3
kullback_leibler_cont_appr(pdf1$y, pdf3$y)

# PDF 2 vs PDF 2
kullback_leibler_cont_appr(pdf2$y, pdf3$y)

oHMMed documentation built on May 29, 2024, 2:20 a.m.