[20] | 1 | /* |
---|
| 2 | * Cloud9: A MapReduce Library for Hadoop |
---|
| 3 | * |
---|
| 4 | * Licensed under the Apache License, Version 2.0 (the "License"); you |
---|
| 5 | * may not use this file except in compliance with the License. You may |
---|
| 6 | * obtain a copy of the License at |
---|
| 7 | * |
---|
| 8 | * http://www.apache.org/licenses/LICENSE-2.0 |
---|
| 9 | * |
---|
| 10 | * Unless required by applicable law or agreed to in writing, software |
---|
| 11 | * distributed under the License is distributed on an "AS IS" BASIS, |
---|
| 12 | * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or |
---|
| 13 | * implied. See the License for the specific language governing |
---|
| 14 | * permissions and limitations under the License. |
---|
| 15 | */ |
---|
| 16 | |
---|
| 17 | package tw.org.nchc.util; |
---|
| 18 | |
---|
| 19 | import java.util.ArrayList; |
---|
| 20 | import java.util.Collections; |
---|
| 21 | import java.util.List; |
---|
| 22 | import java.util.Map; |
---|
| 23 | import java.util.SortedSet; |
---|
| 24 | import java.util.TreeSet; |
---|
| 25 | |
---|
| 26 | /** |
---|
| 27 | * A class for keeping track of the number of times an object has been |
---|
| 28 | * encountered. This is useful for counting things in a stream, e.g., POS tags, |
---|
| 29 | * terms, etc. |
---|
| 30 | */ |
---|
| 31 | public class InstanceCounter<T extends Comparable<T>> { |
---|
| 32 | |
---|
| 33 | // internal representation---although the scores are doubles, counts are |
---|
| 34 | // obviously integers |
---|
| 35 | private ScoreSortedMap<T> mMap; |
---|
| 36 | |
---|
| 37 | private int mTotalCount = 0; |
---|
| 38 | |
---|
| 39 | /** |
---|
| 40 | * Constructs an <code>InstanceCounter</code>. |
---|
| 41 | */ |
---|
| 42 | public InstanceCounter() { |
---|
| 43 | mMap = new ScoreSortedMap<T>(); |
---|
| 44 | } |
---|
| 45 | |
---|
| 46 | /** |
---|
| 47 | * Adds an instance to the set of observations. |
---|
| 48 | * |
---|
| 49 | * @param instance |
---|
| 50 | * the instance observed |
---|
| 51 | */ |
---|
| 52 | public void count(T instance) { |
---|
| 53 | if (mMap.containsKey(instance)) { |
---|
| 54 | mMap.put(instance, mMap.get(instance) + 1); |
---|
| 55 | } else { |
---|
| 56 | mMap.put(instance, 1.0); |
---|
| 57 | } |
---|
| 58 | mTotalCount++; |
---|
| 59 | } |
---|
| 60 | |
---|
| 61 | /** |
---|
| 62 | * Prints each instance and how many times its been observed, sorted by the |
---|
| 63 | * counts. |
---|
| 64 | */ |
---|
| 65 | public void printCounts() { |
---|
| 66 | for (Map.Entry<T, Double> map : mMap.getSortedEntries()) { |
---|
| 67 | System.out.println(map.getValue().intValue() + "\t" + map.getKey()); |
---|
| 68 | } |
---|
| 69 | } |
---|
| 70 | |
---|
| 71 | /** |
---|
| 72 | * Returns a list of <code>InstanceCount</code> objects, sorted by count. |
---|
| 73 | */ |
---|
| 74 | public List<InstanceCount> getCounts() { |
---|
| 75 | List<InstanceCount> l = new ArrayList<InstanceCount>(); |
---|
| 76 | |
---|
| 77 | for (Map.Entry<T, Double> map : mMap.getSortedEntries()) { |
---|
| 78 | l.add(new InstanceCount(map.getKey(), map.getValue().intValue(), |
---|
| 79 | map.getValue() / (double) mTotalCount)); |
---|
| 80 | } |
---|
| 81 | |
---|
| 82 | return Collections.unmodifiableList(l); |
---|
| 83 | } |
---|
| 84 | |
---|
| 85 | /** |
---|
| 86 | * Returns the total number of observations. |
---|
| 87 | * |
---|
| 88 | * @return the total number of observations |
---|
| 89 | */ |
---|
| 90 | public int getTotalCount() { |
---|
| 91 | return mTotalCount; |
---|
| 92 | } |
---|
| 93 | |
---|
| 94 | /** |
---|
| 95 | * Returns the number of times a particular instance has been observed. |
---|
| 96 | * |
---|
| 97 | * @param inst |
---|
| 98 | * the instance |
---|
| 99 | * @return the count of the instance |
---|
| 100 | */ |
---|
| 101 | public int getCount(T inst) { |
---|
| 102 | if (mMap.containsKey(inst)) { |
---|
| 103 | return mMap.get(inst).intValue(); |
---|
| 104 | } |
---|
| 105 | |
---|
| 106 | return 0; |
---|
| 107 | } |
---|
| 108 | |
---|
| 109 | /** |
---|
| 110 | * Returns a collection of all objects observed, sorted by their natural |
---|
| 111 | * order. |
---|
| 112 | * |
---|
| 113 | * @return a collection of all objects observed, sorted by their natural |
---|
| 114 | * order. |
---|
| 115 | */ |
---|
| 116 | public SortedSet<T> getObservedObjects() { |
---|
| 117 | SortedSet<T> t = new TreeSet<T>(); |
---|
| 118 | |
---|
| 119 | for (T obj : mMap.keySet()) { |
---|
| 120 | t.add(obj); |
---|
| 121 | } |
---|
| 122 | |
---|
| 123 | return t; |
---|
| 124 | } |
---|
| 125 | |
---|
| 126 | /** |
---|
| 127 | * A class that holds an instance, its count, and its frequency. |
---|
| 128 | */ |
---|
| 129 | public class InstanceCount { |
---|
| 130 | private T mInstance; |
---|
| 131 | |
---|
| 132 | private int mCount; |
---|
| 133 | |
---|
| 134 | private double mFreq; |
---|
| 135 | |
---|
| 136 | private InstanceCount(T instance, int cnt, double freq) { |
---|
| 137 | mInstance = instance; |
---|
| 138 | mCount = cnt; |
---|
| 139 | mFreq = freq; |
---|
| 140 | } |
---|
| 141 | |
---|
| 142 | /** |
---|
| 143 | * Returns the instance. |
---|
| 144 | */ |
---|
| 145 | public T getInstance() { |
---|
| 146 | return mInstance; |
---|
| 147 | } |
---|
| 148 | |
---|
| 149 | /** |
---|
| 150 | * Returns the number of times the instance has been observed. |
---|
| 151 | */ |
---|
| 152 | public int getCount() { |
---|
| 153 | return mCount; |
---|
| 154 | } |
---|
| 155 | |
---|
| 156 | /** |
---|
| 157 | * Returns the frequency that this instance has been observed. Frequency |
---|
| 158 | * is the count divided by the total number of observed instances. |
---|
| 159 | */ |
---|
| 160 | public double getFrequency() { |
---|
| 161 | return mFreq; |
---|
| 162 | } |
---|
| 163 | } |
---|
| 164 | |
---|
| 165 | public void clear() { |
---|
| 166 | mMap.clear(); |
---|
| 167 | } |
---|
| 168 | |
---|
| 169 | } |
---|