mutualInfo {bioDist} | R Documentation |
Calculate mutual information via binning
mutualInfo(x, nbin = 10, diag = FALSE, upper = FALSE) MIdist(x, nbin = 10, diag = FALSE, upper = FALSE)
x |
an n by p matrix |
nbin |
number of bins to calculate discrete probabilities |
diag |
if TRUE, diagonal of the distance matrix will be displayed |
upper |
if TRUE, upper triangle of the distance matrix will be displayed |
For mutualInfo
each row of x
is divided into
nbin
groups and then the mutual information is computed, treating
the data as if they were discrete.
For MIdist
we use the transformation proposed by Joe (1989),
delta* = (1 - exp(-2 delta))^.5
where delta is the mutual information. The MIdist
is
then 1-delta*. Joe argues that this measure is then
similar to Kendall's tau, tau.dist
.
An object of class dist
which contains the pairwise distances.
Robert Gentleman
dist
, KLdist.matrix
,
cor.dist
, KLD.matrix
x <- matrix(rnorm(100),nrow=5) mutualInfo(x, nbin=3)