Repository files navigation

A package for efficient computations of standard clustering comparisonmeasures

Stable version on theCRAN.

install.packages("aricode")

The development version is available via:

devtools::install_github("jchiquet/aricode")

Computation of measures for clustering comparison (ARI, AMI, NID andeven the (\chi^2) distance) are usually based on the contingencytable. Traditional implementations (e.g., functionadjustedRandIndexof packagemclust) are in (\Omega(n + u v)) where

• (n) is the size of the vectors the classifications of which are tobe compared,
• (u) and (v) are the respective number of classes in eachvectors.

Inaricode we propose an implementation, based on radix sort, thatis in (\Theta(n)) in time and space.
Importantly, the complexity does not depends on (u) and (v). Ourimplementation of the ARI for instance is one or two order of magnitudefaster than some standard implementation inR.

The functions included in aricode are:

• ARI: computes the adjusted rand index
• Chi2: computes the Chi-square statistics
• MARI/MARIraw: computes the modified adjusted rand index (Sundqvistet al, in preparation)
• NVI: computes the the normalized variation information
• NID: computes the normalized information distance
• NMI: computes the normalized mutual information
• AMI: computes the adjusted mutual information
• expected_MI: computes the expected mutual information
• entropy: computes the conditional and joint entropies
• clustComp: computes all clustering comparison measures at once

Here are some timings to compare the cost of computing the adjusted RandIndex witharicode or with the commonly used functionadjustedRandIndex of themclust package: the cost of the latter canbe prohibitive for large vectors: