You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mahout.apache.org by tc...@apache.org on 2012/03/02 20:46:18 UTC
svn commit: r1296412 - in
/mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth:
PFPGrowth.java TransactionTree.java convertors/TransactionIterator.java
fpgrowth2/FPGrowthIds.java fpgrowth2/FPGrowthObj.java fpgrowth2/FPTree.java
Author: tcp
Date: Fri Mar 2 19:46:17 2012
New Revision: 1296412
URL: http://svn.apache.org/viewvc?rev=1296412&view=rev
Log:
a few more cleanups to make Jenkins happier
Modified:
mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/PFPGrowth.java
mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/TransactionTree.java
mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/convertors/TransactionIterator.java
mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthIds.java
mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthObj.java
mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPTree.java
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/PFPGrowth.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/PFPGrowth.java?rev=1296412&r1=1296411&r2=1296412&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/PFPGrowth.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/PFPGrowth.java Fri Mar 2 19:46:17 2012
@@ -179,11 +179,11 @@ public final class PFPGrowth {
return fList;
}
- public static final int getGroup(int itemId, int maxPerGroup) {
+ public static int getGroup(int itemId, int maxPerGroup) {
return itemId / maxPerGroup;
}
- public static final IntArrayList getGroupMembers(int groupId,
+ public static IntArrayList getGroupMembers(int groupId,
int maxPerGroup,
int numFeatures) {
IntArrayList ret = new IntArrayList();
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/TransactionTree.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/TransactionTree.java?rev=1296412&r1=1296411&r2=1296412&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/TransactionTree.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/TransactionTree.java Fri Mar 2 19:46:17 2012
@@ -195,7 +195,6 @@ public final class TransactionTree imple
public TransactionTree getCompressedTree() {
TransactionTree ctree = new TransactionTree();
Iterator<Pair<IntArrayList,Long>> it = iterator();
- final Map<Integer,MutableLong> fList = generateFList();
int node = 0;
int size = 0;
List<Pair<IntArrayList,Long>> compressedTransactionSet = Lists.newArrayList();
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/convertors/TransactionIterator.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/convertors/TransactionIterator.java?rev=1296412&r1=1296411&r2=1296412&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/convertors/TransactionIterator.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/convertors/TransactionIterator.java Fri Mar 2 19:46:17 2012
@@ -43,6 +43,9 @@ public class TransactionIterator<T> exte
@Override
public Pair<int[],Long> apply(Pair<List<T>,Long> from) {
int index = 0;
+ if (from == null) {
+ return null;
+ }
for (T attribute : from.getFirst()) {
if (attributeIdMapping.containsKey(attribute)) {
transactionBuffer[index++] = attributeIdMapping.get(attribute);
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthIds.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthIds.java?rev=1296412&r1=1296411&r2=1296412&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthIds.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthIds.java Fri Mar 2 19:46:17 2012
@@ -19,18 +19,13 @@ package org.apache.mahout.fpm.pfpgrowth.
import java.io.IOException;
import java.util.AbstractMap;
-import java.util.Arrays;
-import java.util.Collection;
-import java.util.Collections;
import java.util.Iterator;
import java.util.List;
import java.util.Map;
-import java.util.Map.Entry;
import java.util.Set;
import com.google.common.collect.Lists;
import com.google.common.collect.Maps;
-import com.google.common.collect.Sets;
import org.apache.commons.lang.mutable.MutableLong;
import org.apache.hadoop.conf.Configuration;
@@ -39,14 +34,11 @@ import org.apache.hadoop.io.Writable;
import org.apache.hadoop.mapred.OutputCollector;
import org.apache.mahout.common.Pair;
import org.apache.mahout.common.iterator.sequencefile.SequenceFileIterable;
-import org.apache.mahout.fpm.pfpgrowth.CountDescendingPairComparator;
import org.apache.mahout.fpm.pfpgrowth.convertors.StatusUpdater;
import org.apache.mahout.fpm.pfpgrowth.convertors.TopKPatternsOutputConverter;
-import org.apache.mahout.fpm.pfpgrowth.convertors.TransactionIterator;
import org.apache.mahout.fpm.pfpgrowth.convertors.string.TopKStringPatterns;
import org.apache.mahout.math.list.LongArrayList;
import org.apache.mahout.math.list.IntArrayList;
-import org.apache.mahout.math.map.OpenIntIntHashMap;
import org.apache.mahout.math.set.OpenIntHashSet;
import org.slf4j.Logger;
@@ -110,7 +102,6 @@ public class FPGrowthIds {
log.info("Number of unique items {}", attributeFrequency.size());
- OpenIntHashSet returnFeatures = new OpenIntHashSet();
if (returnableFeatures == null || returnableFeatures.isEmpty()) {
returnableFeatures = new IntArrayList();
for (int j = 0; j < attributeFrequency.size(); j++) {
@@ -120,7 +111,7 @@ public class FPGrowthIds {
log.info("Number of unique pruned items {}", attributeFrequency.size());
generateTopKFrequentPatterns(transactionStream, attributeFrequency,
- minSupport, k, attributeFrequency.size(), returnableFeatures,
+ minSupport, k, returnableFeatures,
new TopKPatternsOutputConverter<Integer>(output, new IdentityMapping()), updater);
}
@@ -206,7 +197,6 @@ public class FPGrowthIds {
LongArrayList attributeFrequency,
long minSupport,
int k,
- int featureSetSize,
IntArrayList returnFeatures,
TopKPatternsOutputConverter<Integer> topKPatternsOutputCollector,
StatusUpdater updater) throws IOException {
@@ -271,8 +261,7 @@ public class FPGrowthIds {
}
if (prefixPats != null) {
- FrequentPatternMaxHeap toRet = cross(prefixPats, suffixPats, k);
- return toRet;
+ return cross(prefixPats, suffixPats, k);
}
return suffixPats;
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthObj.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthObj.java?rev=1296412&r1=1296411&r2=1296412&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthObj.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPGrowthObj.java Fri Mar 2 19:46:17 2012
@@ -18,7 +18,6 @@
package org.apache.mahout.fpm.pfpgrowth.fpgrowth2;
import java.io.IOException;
-import java.util.Arrays;
import java.util.Collection;
import java.util.Collections;
import java.util.Iterator;
@@ -42,7 +41,6 @@ import org.apache.mahout.fpm.pfpgrowth.c
import org.apache.mahout.fpm.pfpgrowth.convertors.TopKPatternsOutputConverter;
import org.apache.mahout.fpm.pfpgrowth.convertors.TransactionIterator;
import org.apache.mahout.fpm.pfpgrowth.convertors.string.TopKStringPatterns;
-import org.apache.mahout.math.map.OpenIntIntHashMap;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
@@ -176,8 +174,8 @@ public class FPGrowthObj<A extends Compa
log.info("Number of unique pruned items {}", attributeIdMapping.size());
generateTopKFrequentPatterns(new TransactionIterator<A>(transactionStream,
- attributeIdMapping), attributeFrequency, minSupport, k, reverseMapping
- .size(), returnFeatures, new TopKPatternsOutputConverter<A>(output,
+ attributeIdMapping), attributeFrequency, minSupport, k,
+ returnFeatures, new TopKPatternsOutputConverter<A>(output,
reverseMapping), updater);
}
@@ -250,7 +248,7 @@ public class FPGrowthObj<A extends Compa
long[] attributeFrequency,
long minSupport,
int k,
- int featureSetSize,
+ //int featureSetSize,
Collection<Integer> returnFeatures, TopKPatternsOutputConverter<A> topKPatternsOutputCollector,
StatusUpdater updater) throws IOException {
@@ -317,8 +315,7 @@ public class FPGrowthObj<A extends Compa
}
if (prefixPats != null) {
- FrequentPatternMaxHeap toRet = cross(prefixPats, suffixPats, k);
- return toRet;
+ return cross(prefixPats, suffixPats, k);
}
return suffixPats;
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPTree.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPTree.java?rev=1296412&r1=1296411&r2=1296412&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPTree.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/fpgrowth2/FPTree.java Fri Mar 2 19:46:17 2012
@@ -113,8 +113,8 @@ public class FPTree {
this.attrCountList = new LongArrayList();
for (int i = 0; i < attrCounts.length; i++)
if (attrCounts[i] > 0) {
- if (attrCountList.size() < (i+1)) {
- attrCountList.setSize(i+1);
+ if (attrCountList.size() < (i + 1)) {
+ attrCountList.setSize(i + 1);
}
attrCountList.set(i, attrCounts[i]);
}
@@ -248,7 +248,7 @@ public class FPTree {
while (currNode != root) {
int currAttr = currNode.attribute();
if (counts.size() <= currAttr) {
- counts.setSize(currAttr+1);
+ counts.setSize(currAttr + 1);
}
long count = counts.get(currAttr);
counts.set(currNode.attribute(), count + pathCount);
@@ -257,9 +257,9 @@ public class FPTree {
}
if (counts.get(targetAttr) != attrCountList.get(targetAttr))
throw new IllegalStateException("mismatched counts for targetAttr="
- +targetAttr+", ("+counts.get(targetAttr)
- +" != "+attrCountList.get(targetAttr)+"); "
- +"thisTree="+this+"\n");
+ + targetAttr + ", (" + counts.get(targetAttr)
+ + " != " + attrCountList.get(targetAttr) + "); "
+ + "thisTree=" + this + "\n");
counts.set(targetAttr, 0L);
FPTree toRet = new FPTree(counts, minSupport);
@@ -280,7 +280,7 @@ public class FPTree {
}
// biggest count or smallest attr number goes first
- private class AttrComparator implements Comparator<Integer>{
+ private class AttrComparator implements Comparator<Integer> {
public int compare(Integer a, Integer b) {
long aCnt = 0;
@@ -310,7 +310,7 @@ public class FPTree {
while (currNode.numChildren() == 1) {
currNode = currNode.children().iterator().next();
if (pAttrCountList.size() <= currNode.attribute())
- pAttrCountList.setSize(currNode.attribute()+1);
+ pAttrCountList.setSize(currNode.attribute() + 1);
pAttrCountList.set(currNode.attribute(), currNode.count());
qAttrCountList.set(currNode.attribute(), 0);
}
@@ -336,13 +336,13 @@ public class FPTree {
items.add(attribute);
}
for (FPNode child : node.children()) {
- added+= recursivelyAddPrefixPats(pTree, qTree, child, items);
+ added += recursivelyAddPrefixPats(pTree, qTree, child, items);
}
if (added < count) {
long toAdd = count - added;
pTree.accumulate(items, toAdd);
qTree.accumulate(items, toAdd);
- added+= toAdd;
+ added += toAdd;
}
if (!(node == root)) {
int lastIdx = items.size() - 1;