-
Notifications
You must be signed in to change notification settings - Fork 89
/
FisherDiscriminant.java
165 lines (139 loc) · 5.33 KB
/
FisherDiscriminant.java
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
/*
* avenir: Predictive analytic based on Hadoop Map Reduce
* Author: Pranab Ghosh
*
* Licensed 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.avenir.discriminant;
import java.io.IOException;
import java.util.HashMap;
import java.util.Map;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
import org.chombo.mr.NumericalAttrStats;
import org.chombo.util.Tuple;
import org.chombo.util.Utility;
/**
* Fisher univariate discriminant. Feature is numeric. Classification is binary
* @author pranab
*
*/
public class FisherDiscriminant extends Configured implements Tool {
@Override
public int run(String[] args) throws Exception {
Job job = new Job(getConf());
String jobName = "Univariate Fisher linear discriminant";
job.setJobName(jobName);
job.setJarByClass(FisherDiscriminant.class);
FileInputFormat.addInputPath(job, new Path(args[0]));
FileOutputFormat.setOutputPath(job, new Path(args[1]));
Utility.setConfiguration(job.getConfiguration(), "avenir");
job.setMapperClass(NumericalAttrStats.StatsMapper.class);
job.setReducerClass(FisherDiscriminant.FisherReducer.class);
job.setCombinerClass(NumericalAttrStats.StatsCombiner.class);
job.setMapOutputKeyClass(Tuple.class);
job.setMapOutputValueClass(Tuple.class);
job.setOutputKeyClass(NullWritable.class);
job.setOutputValueClass(Text.class);
job.setNumReduceTasks(job.getConfiguration().getInt("num.reducer", 1));
int status = job.waitForCompletion(true) ? 0 : 1;
return status;
}
/**
* @author pranab
*
*/
public static class FisherReducer extends NumericalAttrStats.StatsReducer {
private static final String uncondAttrVal = "0";
private Map<Integer, ConditionedFeatureStat[]> attrCondStats = new HashMap<Integer, ConditionedFeatureStat[]>();
/* (non-Javadoc)
* @see org.apache.hadoop.mapreduce.Reducer#cleanup(org.apache.hadoop.mapreduce.Reducer.Context)
*/
protected void cleanup(Context context) throws IOException, InterruptedException {
//emit class boundary
for (int attr : attrCondStats.keySet()) {
ConditionedFeatureStat[] condStats = attrCondStats.get(attr);
double pooledVariance = (condStats[0].getVariance() * condStats[0].getCount() +
condStats[1].getVariance() * condStats[1].getCount()) / (condStats[0].getCount() + condStats[1].getCount());
double logOddsPrior = Math.log((double)condStats[0].getCount() / condStats[1].getCount());
double meanDiff = condStats[0].getMean() - condStats[1].getMean();
double discrimValue = (condStats[0].getMean() + condStats[1].getMean()) / 2;
discrimValue -= logOddsPrior * pooledVariance / meanDiff;
outVal.set("" + attr + fieldDelim + logOddsPrior + fieldDelim + pooledVariance + fieldDelim + discrimValue);
context.write(NullWritable.get(), outVal);
}
}
protected void reduce(Tuple key, Iterable<Tuple> values, Context context)
throws IOException, InterruptedException {
processReduce(values);
//process conditional stats
String condAttrVal = key.getString(1);
if (!uncondAttrVal.equals(condAttrVal)) {
Integer attr = key.getInt(0);
ConditionedFeatureStat[] condStats = attrCondStats.get(attr);
if (null == condStats) {
condStats = new ConditionedFeatureStat[2];
condStats[0] = condStats[1] = null;
attrCondStats.put(attr, condStats);
}
int indx = condStats[0] == null ? 0 : 1;
condStats[indx] = new ConditionedFeatureStat(condAttrVal, totalCount, mean, variance);
}
//emit conditional stat
emitOutput( key, context);
}
}
/**
* @author pranab
*
*/
public static class ConditionedFeatureStat {
private String condAttrVal;
private int count;
private double mean;
private double variance;
public ConditionedFeatureStat(String condAttrVal, int count, double mean, double variance) {
super();
this.condAttrVal = condAttrVal;
this.count = count;
this.mean = mean;
this.variance = variance;
}
public String getCondAttrVal() {
return condAttrVal;
}
public int getCount() {
return count;
}
public double getMean() {
return mean;
}
public double getVariance() {
return variance;
}
}
/**
* @param args
*/
public static void main(String[] args) throws Exception {
int exitCode = ToolRunner.run(new FisherDiscriminant(), args);
System.exit(exitCode);
}
}