/ 
String-Cluster-Hobohm-0.141020
River stage zero No dependents
/ String::Cluster::Hobohm

NAME

String::Cluster::Hobohm - Cluster strings using the Hobohm algorithm

VERSION

version 0.141020

SYNOPSIS

use String::Cluster::Hobohm;

my @strings = qw(foo foa bar);

my $clusterer = String::Cluster::Hobohm->new( similarity => 0.62 );

my $groups = $clusterer->cluster( \@strings );

# [ [ \'foo', \'foa' ], [ \'bar' ] ];

my @reduced = map { ${ $_->[0] } } @$groups;

# [ 'foo', 'bar' ];

DESCRIPTION

String::Cluster::Hobohm implements the Hobohm clustering algorithm [1], originally devised to reduce redundancy of biological sequence data sets.

As a clustering algorithm, it takes a set of sequences, and returns them grouped by similarity. The latter is computed using the Levenshtein distance, as implemented by the Text::LevenshteinXS module.

ATTRIBUTES

similarity

The similarity threshold that defines whether two strings are sufficiently alike to be part of the same cluster. Should be a number between 0 and 1. Defaults to 0.62.

METHODS

cluster

my $grouped = $hobohm->cluster( \@strings );

Takes an array reference with the sequences to cluster as argument, and returns an array reference of clusters. Each cluster is depicted as a list of references to the strings that define it.

For example, given the following array of strings, and a similarity of 0.62:

[ 'foo', 'foa', 'bar' ];

The data structure returned after clustering would be:

[ [ \'foo', \'foa' ], [ \'bar' ] ];

The reason for using references instead of the actual strings is to avoid copying potentially large strings and taking up too much memory (remember that the algorithm was designed with biological sequences in mind).

REFERENCES

[1] Uwe Hobohm, Michael Scharf, Reinhard Schneider and Chris Sander. Selection of representative protein data sets. Protein Science (1992), 409-417. Cambridge University Press.

AUTHOR

Bruno Vecchi <vecchi.b gmail.com>

COPYRIGHT AND LICENSE

This software is copyright (c) 2014 by Bruno Vecchi.

This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.