1. 用Maven创建一个标准化的Java项目
2. 导入项目到eclipse
3. 增加mahout依赖，修改pom.xml
4. 下载依赖

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
<modelVersion>4.0.0</modelVersion>
<groupId>org.conan.mymahout</groupId>
<artifactId>myMahout</artifactId>
<packaging>jar</packaging>
<version>1.0-SNAPSHOT</version>
<name>myMahout</name>
<url>http://maven.apache.org</url>

<properties>
<project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
<mahout.version>0.6</mahout.version>
</properties>

<dependencies>
<dependency>
<groupId>org.apache.mahout</groupId>
<artifactId>mahout-core</artifactId>
<version>${mahout.version}</version>
</dependency>
<dependency>
<groupId>org.apache.mahout</groupId>
<artifactId>mahout-integration</artifactId>
<version>${mahout.version}</version>
<exclusions>
<exclusion>
<groupId>org.mortbay.jetty</groupId>
<artifactId>jetty</artifactId>
</exclusion>
<exclusion>
<groupId>org.apache.cassandra</groupId>
<artifactId>cassandra-all</artifactId>
</exclusion>
<exclusion>
<groupId>me.prettyprint</groupId>
<artifactId>hector-core</artifactId>
</exclusion>
</exclusions>
</dependency>
</dependencies>
</project>

~ mvn clean install

•  准备数据文件: item.csv
•  Java程序：UserCF.java
•  运行程序
•  推荐结果解读

package org.conan.mymahout.recommendation;

import java.io.File;
import java.io.IOException;
import java.util.List;

import org.apache.mahout.cf.taste.common.TasteException;
import org.apache.mahout.cf.taste.impl.common.LongPrimitiveIterator;
import org.apache.mahout.cf.taste.impl.model.file.FileDataModel;
import org.apache.mahout.cf.taste.impl.neighborhood.NearestNUserNeighborhood;
import org.apache.mahout.cf.taste.impl.recommender.GenericUserBasedRecommender;
import org.apache.mahout.cf.taste.impl.similarity.EuclideanDistanceSimilarity;
import org.apache.mahout.cf.taste.model.DataModel;
import org.apache.mahout.cf.taste.recommender.RecommendedItem;
import org.apache.mahout.cf.taste.recommender.Recommender;
import org.apache.mahout.cf.taste.similarity.UserSimilarity;

public class UserCF {

final static int NEIGHBORHOOD_NUM = 2;
final static int RECOMMENDER_NUM = 3;

public static void main(String[] args) throws IOException, TasteException {
String file = "datafile/item.csv";
DataModel model = new FileDataModel(new File(file));
UserSimilarity user = new EuclideanDistanceSimilarity(model);
NearestNUserNeighborhood neighbor = new NearestNUserNeighborhood(NEIGHBORHOOD_NUM, user, model);
Recommender r = new GenericUserBasedRecommender(model, neighbor, user);
LongPrimitiveIterator iter = model.getUserIDs();

while (iter.hasNext()) {
long uid = iter.nextLong();
List list = r.recommend(uid, RECOMMENDER_NUM);
System.out.printf("uid:%s", uid);
for (RecommendedItem ritem : list) {
System.out.printf("(%s,%f)", ritem.getItemID(), ritem.getValue());
}
System.out.println();
}
}
}

3). 运行程序

SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinder".
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#StaticLoggerBinder for further details.
uid:1(104,4.274336)(106,4.000000)
uid:2(105,4.055916)
uid:3(103,3.360987)(102,2.773169)
uid:4(102,3.000000)
uid:5

~ vi datafile/randomData.csv

-0.883033363823402,-3.31967192630249
-2.39312626419456,3.34726861118871
2.66976353341256,1.85144276077058
-1.09922906899594,-6.06261735207489
-4.36361936997216,1.90509905380532
-0.00351835125495037,-0.610105996559153
-2.9962958796338,-3.60959839525735
-3.27529418132066,0.0230099799641799
2.17665594420569,6.77290756817957
-2.47862038335637,2.53431833167278
5.53654901906814,2.65089785582474
5.66257474538338,6.86783609641077
-0.558946883114376,1.22332819416237
5.11728525486132,3.74663871584768
1.91240516693351,2.95874731384062
-2.49747101306535,2.05006504756875
3.98781883213459,1.00780938946366

这里只截取了一部分，更多的数据请查看源代码。

注：我是通过R语言生成的randomData.csv

x1<-cbind(x=rnorm(400,1,3),y=rnorm(400,1,3))
x2<-cbind(x=rnorm(300,1,0.5),y=rnorm(300,0,0.5))
x3<-cbind(x=rnorm(300,0,0.1),y=rnorm(300,2,0.2))
x<-rbind(x1,x2,x3)
write.table(x,file="randomData.csv",sep=",",row.names=FALSE,col.names=FALSE)

2). Java程序：Kmeans.java

Mahout中kmeans方法的算法实现过程。

新建JAVA类：org.conan.mymahout.cluster06.Kmeans.java

package org.conan.mymahout.cluster06;

import java.io.IOException;
import java.util.ArrayList;
import java.util.List;

import org.apache.mahout.clustering.kmeans.Cluster;
import org.apache.mahout.clustering.kmeans.KMeansClusterer;
import org.apache.mahout.common.distance.EuclideanDistanceMeasure;
import org.apache.mahout.math.Vector;

public class Kmeans {

public static void main(String[] args) throws IOException {
List sampleData = MathUtil.readFileToVector("datafile/randomData.csv");

int k = 3;
double threshold = 0.01;

List randomPoints = MathUtil.chooseRandomPoints(sampleData, k);
for (Vector vector : randomPoints) {
System.out.println("Init Point center: " + vector);
}

List clusters = new ArrayList();
for (int i = 0; i < k; i++) {
clusters.add(new Cluster(randomPoints.get(i), i, new EuclideanDistanceMeasure()));
}

List<List> finalClusters = KMeansClusterer.clusterPoints(sampleData, clusters, new EuclideanDistanceMeasure(), k, threshold);
for (Cluster cluster : finalClusters.get(finalClusters.size() - 1)) {
System.out.println("Cluster id: " + cluster.getId() + " center: " + cluster.getCenter().asFormatString());
}
}

}

3). 运行Java程序

控制台输出:

Init Point center: {0:-0.162693685149196,1:2.19951550286862}
Init Point center: {0:-0.0409782183083317,1:2.09376666042057}
Init Point center: {0:0.158401778474687,1:2.37208412905273}
SLF4J: Failed to load class "org.slf4j.impl.StaticLoggerBinder".
SLF4J: Defaulting to no-operation (NOP) logger implementation
SLF4J: See http://www.slf4j.org/codes.html#StaticLoggerBinder for further details.
Cluster id: 0 center: {0:-2.686856800552941,1:1.8939462954763795}
Cluster id: 1 center: {0:0.6334255423230666,1:0.49472852972602105}
Cluster id: 2 center: {0:3.334520309711998,1:3.2758355898247653}

4). mahout结果解读

1. Init Point center表示，kmeans算法初始时的设置的3个中心点

2. Cluster center表示，聚类后找到3个中心点

5). 用R语言实现Kmeans算法

接下来为了让结果更直观，我们再用R语言，进行kmeans实验，操作相同的数据。

R语言代码：

> y<-read.csv(file="randomData.csv",sep=",",header=FALSE) 
> cl<-kmeans(y,3,iter.max = 10, nstart = 25) 
> cl$centers
V1 V2
1 -0.4323971 2.2852949
2 0.9023786 -0.7011153
3 4.3725463 2.4622609

# 生成聚类中心的图形
> plot(y, col=c("black","blue","green")[cl$cluster])
> points(cl$centers, col="red", pch = 19)

# 画出Mahout聚类的中心
> mahout<-matrix(c(-2.686856800552941,1.8939462954763795,0.6334255423230666,0.49472852972602105,3.334520309711998,3.2758355898247653),ncol=2,byrow=TRUE) 
> points(mahout, col="violetred", pch = 19)

聚类的效果图：

6). 比较Mahout和R的结果

从上图中，我们看到有 黑，蓝，绿，三种颜色的空心点，这些点就是原始的数据。

3个红色实点，是R语言kmeans后生成的3个中心。

3个紫色实点，是Mahout的kmeans后生成的3个中心。

R语言和Mahout生成的点，并不是重合的，原因有几点：

• 距离算法不一样：
•       Mahout中，我们用的 “欧氏距离(EuclideanDistanceMeasure)”
•       R语言中，默认是”Hartigan and Wong”
• 初始化的中心是不一样的。
• 最大迭代次数是不一样的。
• 点合并时，判断的”阈值(threshold)”是不一样的。