startup

Class Similarity

java.lang.Object

startup.Similarity

public class Similarity
extends Object

A [TODO mutable?] measure of similarity between multi-interval sets of Strings.

An instance of Similarity uses a client-provided definition of label similarities, where 0 is least similar and 1 is most similar.

Given two multi-interval sets, let min be the minimum start of any of their intervals, and let max be the maximum end. The similarity between the two sets is the ratio: (sum of piecewise-matching between the sets) / (max - min)

The amount of piecewise-matching for any unit interval [i, i+1) is:

• 0 if neither set has a label on that interval
• 0 if only one set has a label on that interval
• otherwise, the similarity between the labels as defined for this Similarity instance

For example, suppose you have multi-interval sets that use labels "happy", "sad", and "meh"; and similarity between labels is defined as:

• 1 if both are "happy", both "sad", or both "meh"
• 0.5 if one is "meh" and the other is "happy" or "sad"
• 0 otherwise

Then the similarity between these two sets:

• { "happy" = [[0, 1), [2,4)], "sad" = [[1,2)] }
• { "sad" = [[1, 2)], "meh" = [[2,3)], "happy" = [[3,4)] }

would be: (0 + 1 + 0.5 + 1) / (4 - 0) = 0.625

PS2 instructions: this is a required ADT class, and you MUST NOT weaken the required specifications. However you MAY strengthen the specifications and you MAY add additional methods.

Constructor Summary

Constructors

Constructor and Description
Similarity(File similarities) Create a new Similarity where similarity between labels is defined in the given file.

Method Summary

Modifier and Type	Method and Description
double	similarity(MultiIntervalSet<String> a, MultiIntervalSet<String> b) Compute similarity between two multi-interval sets.

Methods inherited from class java.lang.Object

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

Constructor Detail

Similarity

public Similarity(File similarities)
           throws IOException

Create a new Similarity where similarity between labels is defined in the given file. Each line of similarities must contain exactly three pieces, separated by one or more spaces. The first two pieces give a pair of strings, and the third piece gives the decimal similarity between them, in a format allowed by Double.valueOf(String), between 0 and 1 inclusive. Similarity between labels is symmetric, so the order of strings in the pair is irrelevant. A pair may not appear more than once. The similarity between all other pairs of strings is 0. This format cannot define non-zero similarity for strings that contain newlines or spaces, or for the empty string.

For example, the following file defines the similarity function used in the example at the top of this class:

 happy happy 1
 sad   sad   1
 meh   meh   1
 meh   happy 0.5
 meh   sad   0.5
 

Parameters:

similarities - label similarity definition as described above

Throws:

IOException - if the similarity file cannot be found or read

Method Detail

similarity

public double similarity(MultiIntervalSet<String> a,
                         MultiIntervalSet<String> b)

Compute similarity between two multi-interval sets. Returns a value between 0 and 1 inclusive.

Parameters:

a - non-empty multi-interval set of strings

b - non-empty multi-interval set of strings

Returns:

similarity between a and b as defined above