org.apache.mahout.common.distance.TanimotoDistanceMeasure Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of mahout-mr Show documentation

Scalable machine learning libraries

There is a newer version: 0.13.0

/**
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package org.apache.mahout.common.distance;

import org.apache.mahout.math.Vector;
import org.apache.mahout.math.function.Functions;

/**
 * Tanimoto coefficient implementation.
 * 
 * http://en.wikipedia.org/wiki/Jaccard_index
 */
public class TanimotoDistanceMeasure extends WeightedDistanceMeasure {
  
  /**
   * Calculates the distance between two vectors.
   * 
   * The coefficient (a measure of similarity) is: T(a, b) = a.b / (|a|^2 + |b|^2 - a.b)
   * 
   * The distance d(a,b) = 1 - T(a,b)
   * 
   * @return 0 for perfect match, > 0 for greater distance
   */
  @Override
  public double distance(Vector a, Vector b) {
    double ab;
    double denominator;
    if (getWeights() != null) {
      ab = a.times(b).aggregate(getWeights(), Functions.PLUS, Functions.MULT);
      denominator = a.aggregate(getWeights(), Functions.PLUS, Functions.MULT_SQUARE_LEFT)
          + b.aggregate(getWeights(), Functions.PLUS, Functions.MULT_SQUARE_LEFT)
          - ab;
    } else {
      ab = b.dot(a); // b is SequentialAccess
      denominator = a.getLengthSquared() + b.getLengthSquared() - ab;
    }
    
    if (denominator < ab) { // correct for fp round-off: distance >= 0
      denominator = ab;
    }
    if (denominator > 0) {
      // denominator == 0 only when dot(a,a) == dot(b,b) == dot(a,b) == 0
      return 1.0 - ab / denominator;
    } else {
      return 0.0;
    }
  }

  @Override
  public double distance(double centroidLengthSquare, Vector centroid, Vector v) {
    return distance(centroid, v); // TODO
  }
  
}