entropy: Information entropy

In AkselA/R-ymse: Ymse (Various)

Description

Computes the information entropy (also called Shannon entropy) of a set of discrete values, or a tabulated such set.

Usage

entropy(x, ...)

## S3 method for class 'table'
entropy(x, base = 2, ...)

## S3 method for class 'data.frame'
entropy(x, base = 2, ...)

## S3 method for class 'matrix'
entropy(x, base = 2, ...)

## Default S3 method:
entropy(x, base = 2, ...)

Arguments

x	a vector, table, data.frame or matrix. In the case of table, data.frame and matrix each row is treated as a separate set of counts or proportions, with columns representing species, types, categories etc.
...	further arguments passed to methods
base	the log base to be used.

Examples

entropy(c(5, 5, 4, 4, 2, 3, 5))  # default is unit bits
entropy(c(5, 5, 4, 4, 2, 3, 5), base=exp(1))  # unit nats

entropy(rep(1:4, 1:4), 4)
entropy(rep(1:4, 1), 4)

entropy(as.factor(c(1, 1, 2, 3, 4, 4)))
entropy(as.character(c(1, 1, 2, 3, 4, 4)))

mtctab <- table(mtcars$cyl, mtcars$carb)
entropy(mtctab, 6)

xx <- data.frame(bee=c(0, 0, 1, 2, 3, 2, 0, 3),
                wasp=c(1, 3, 2, 0, 1, 1, 2, 1),
                  fly=c(1, 2, 4, 2, 1, 0, 1, 0),
              beetle=c(1, 0, 0, 1, 2, 2, 0, 2),
            butterfly=c(0, 0, 0, 0, 3, 1, 0, 1))

entropy(xx) 

AkselA/R-ymse documentation built on March 21, 2020, 9:52 a.m.