@Override
public void reduce(
IntWritable key,
Iterator<Text> values,
OutputCollector<IntWritable, Text> output,
Reporter reporter)
throws IOException {
StringBuilder sb = new StringBuilder();
while (values.hasNext()) {
sb.append("," + values.next());
}
v.set(sb.toString().replaceFirst(",", "")); // 去掉第一个逗号
output.collect(key, v);
// 键为itemID，值为其打过分的所有用户的偏好
System.out.println(key.toString() + " " + v.toString());
}
@Override
protected void map(Centroid key, IntWritable value, Context context)
throws IOException, InterruptedException {
context.write(new Text(key.toString()), new Text(value.toString()));
}
@Override
public String toString() {
return "<key: " + key.toString() + ", partition: " + partition.toString() + ">";
}
@Override
public void reduce(
IntWritable key,
Iterator<ClusterWritable> values,
OutputCollector<IntWritable, Text> output,
Reporter reporter)
throws IOException {
float sumSimilarity = 0.0f;
int numMovies = 0;
float avgSimilarity = 0.0f;
float similarity = 0.0f;
int s = 0;
int count;
float diff = 0.0f;
float minDiff = 1.0f;
int candidate = 0;
String data = new String("");
String shortline = new String("");
ArrayList<String> arrl = new ArrayList<String>();
ArrayList<Float> simArrl = new ArrayList<Float>();
String oneElm = new String();
int indexShort, index2;
Text val = new Text();
while (values.hasNext()) {
ClusterWritable cr = (ClusterWritable) values.next();
similarity = cr.similarity;
simArrl.addAll(cr.similarities);
for (int i = 0; i < cr.movies.size(); i++) {
oneElm = cr.movies.get(i);
indexShort =
oneElm.indexOf(
",",
1000); // to avoid memory error caused by long arrays; it will results less
// accurate
if (indexShort == -1) {
shortline = new String(oneElm);
} else {
shortline = new String(oneElm.substring(0, indexShort));
}
arrl.add(shortline);
output.collect(key, new Text(oneElm));
}
numMovies += cr.movies.size();
sumSimilarity += similarity;
}
if (numMovies > 0) {
avgSimilarity = sumSimilarity / (float) numMovies;
}
diff = 0.0f;
minDiff = 1.0f;
for (s = 0; s < numMovies; s++) {
diff = (float) Math.abs(avgSimilarity - simArrl.get(s));
if (diff < minDiff) {
minDiff = diff;
candidate = s;
}
}
data = arrl.get(candidate);
index2 = data.indexOf(":");
String movieStr = data.substring(0, index2);
String reviews = data.substring(index2 + 1);
StringTokenizer token = new StringTokenizer(reviews, ",");
count = 0;
while (token.hasMoreTokens()) {
token.nextToken();
count++;
}
System.out.println(
"The key = "
+ key.toString()
+ " has members = "
+ numMovies
+ " simil = "
+ simArrl.get(candidate));
val = new Text(simArrl.get(candidate) + " " + movieStr + " " + count + " " + reviews);
output.collect(key, val);
reporter.incrCounter(Counter.VALUES, 1);
}
@Override
public String toString() {
return first.toString() + ":" + second.toString();
}
