org.lsst.ccs.utilities.misc

Class Similarity

java.lang.Object

org.lsst.ccs.utilities.misc.Similarity

public class Similarity
extends Object

Tests strings for similarity based on Levenshtein distance.

Author:

onoprien

Constructor Summary

Constructors

Constructor and Description
Similarity()

Method Summary

Modifier and Type	Method and Description
static int	compute(String s1, String s2) Computes similarity score for two strings using Levenshtein distance.
static int	computeBiased(String probe, String candidate) Computes similarity score for two strings using modified Levenshtein distance.
static ArrayList<String>	head(String probe, List<String> candidates, int max, int threshold) Returns a list of strings from candidates that are most similar to probe.
static ArrayList<String>	head(String probe, List<String> candidates, int max, int threshold, BiFunction<String,String,Integer> computeScore) Returns a list of strings from candidates that are most similar to probe.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Similarity

public Similarity()

Method Detail

compute

public static int compute(String s1,
                          String s2)

Computes similarity score for two strings using Levenshtein distance. The lower the score, the more similar strings are. Identical strings yield score of 0.

Parameters:

s1 - First string to compare.

s2 - Second string to compare.

Returns:

Similarity score.

computeBiased

public static int computeBiased(String probe,
                                String candidate)

Computes similarity score for two strings using modified Levenshtein distance. If the candidate is longer than the probe, its beginning is used for the comparison. The lower the score, the more similar strings are. Identical strings yield score of 0.

Parameters:

probe - probe string.

candidate - Candidate string

Returns:

Similarity score.

head

public static ArrayList<String> head(String probe,
                                     List<String> candidates,
                                     int max,
                                     int threshold)

Returns a list of strings from candidates that are most similar to probe. Similarity is computed using Levenshtein distance.

Parameters:

probe - Test string.

candidates - List of strings from which those most similar to probe should be selected.

max - Maximum number of strings to select.

threshold - Only strings with similarity score below threshold can be selected.

Returns:

List of selected strings.

head

public static ArrayList<String> head(String probe,
                                     List<String> candidates,
                                     int max,
                                     int threshold,
                                     BiFunction<String,String,Integer> computeScore)

Returns a list of strings from candidates that are most similar to probe.

Parameters:

probe - Test string.

candidates - List of strings from which those most similar to probe should be selected.

max - Maximum number of strings to select.

threshold - Only strings with similarity score below threshold can be selected.

computeScore - Function that computes similarity score between probe and candidate.

Returns:

List of selected strings.