Small additional improvements for AUC

felixtang · Nov 18, 2012 · 884fc53 · 884fc53
1 parent 7218ad9
commit 884fc53
Show file tree

Hide file tree

Showing 4 changed files with 19 additions and 28 deletions.
diff --git a/CHANGES.html b/CHANGES.html
@@ -14,6 +14,12 @@
 
 <h1>Version 0.9</h1>
 
+<h2>Other</h2>
+
+<ul>
+  <li>Faster display of self-organizing map visualization for moderate to large data sets.</li>
+</ul>
+
 <h1>Version 0.8</h1>
 
 <p>November 16 2012</p>

diff --git a/common/src/net/myrrix/common/ByValueAscComparator.java b/common/src/net/myrrix/common/ByValueAscComparator.java
@@ -26,9 +26,9 @@
  *
  * @author Sean Owen
  */
-final class ByValueAscComparator implements Comparator<RecommendedItem> {
+public final class ByValueAscComparator implements Comparator<RecommendedItem> {
 
-  static final Comparator<RecommendedItem> INSTANCE = new ByValueAscComparator();
+  public static final Comparator<RecommendedItem> INSTANCE = new ByValueAscComparator();
 
   private ByValueAscComparator() {
   }

diff --git a/online/src/net/myrrix/online/eval/AUCEvaluator.java b/online/src/net/myrrix/online/eval/AUCEvaluator.java
@@ -24,7 +24,6 @@
 import org.apache.mahout.cf.taste.common.NoSuchItemException;
 import org.apache.mahout.cf.taste.common.NoSuchUserException;
 import org.apache.mahout.cf.taste.common.TasteException;
-import org.apache.mahout.cf.taste.impl.common.LongPrimitiveIterator;
 import org.apache.mahout.cf.taste.recommender.RecommendedItem;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
@@ -56,17 +55,19 @@ public EvaluationResult evaluate(MyrrixRecommender recommender,
     int underCurve = 0;
     int total = 0;
 
-    long[] allItemIDs = getAllItemIDs(recommender);
+    long[] allItemIDs = recommender.getAllItemIDs().toArray();
     RandomGenerator random = RandomManager.getRandom();
 
+    FastIDSet testItemIDs = new FastIDSet();
+
     for (long userID : testData.keySet()) {
 
       List<RecommendedItem> testValues = Lists.newArrayList(testData.get(userID));
       int numTest = testValues.size();
       if (numTest == 0) {
         continue;
       }
-      FastIDSet testItemIDs = new FastIDSet();
+      testItemIDs.clear();
       for (RecommendedItem testValue : testValues) {
         testItemIDs.add(testValue.getItemID());
       }
@@ -98,7 +99,7 @@ public EvaluationResult evaluate(MyrrixRecommender recommender,
         total++;
       }
 
-      if (++count % 10000 == 0) {
+      if (++count % 100000 == 0) {
         log.info("AUC: {}", (double) underCurve / total);
       }
     }
@@ -108,15 +109,4 @@ public EvaluationResult evaluate(MyrrixRecommender recommender,
     return new EvaluationResultImpl(score);
   }
 
-  private static long[] getAllItemIDs(MyrrixRecommender recommender) throws TasteException {
-    FastIDSet allItemIDsSet = recommender.getAllItemIDs();
-    long[] allItemIDs = new long[allItemIDsSet.size()];
-    LongPrimitiveIterator it = allItemIDsSet.iterator();
-    int i = 0;
-    while (it.hasNext()) {
-      allItemIDs[i++] = it.nextLong();
-    }
-    return allItemIDs;
-  }
-
 }
diff --git a/online/src/net/myrrix/online/eval/AbstractEvaluator.java b/online/src/net/myrrix/online/eval/AbstractEvaluator.java
@@ -16,18 +16,17 @@
 
 package net.myrrix.online.eval;
 
-import java.io.BufferedOutputStream;
 import java.io.File;
 import java.io.FileOutputStream;
 import java.io.IOException;
 import java.io.OutputStreamWriter;
 import java.io.Writer;
 import java.util.Collection;
 import java.util.Collections;
-import java.util.Comparator;
 import java.util.Iterator;
 import java.util.List;
 import java.util.Map;
+import java.util.zip.GZIPOutputStream;
 
 import com.google.common.base.CharMatcher;
 import com.google.common.base.Charsets;
@@ -47,6 +46,7 @@
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
+import net.myrrix.common.ByValueAscComparator;
 import net.myrrix.common.io.IOUtils;
 import net.myrrix.common.LangUtils;
 import net.myrrix.common.MyrrixRecommender;
@@ -101,7 +101,7 @@ public final EvaluationResult evaluate(File originalDataDir,
 
     File trainingDataDir = Files.createTempDir();
     trainingDataDir.deleteOnExit();
-    File trainingFile = new File(trainingDataDir, "training.csv");
+    File trainingFile = new File(trainingDataDir, "training.csv.gz");
     trainingFile.deleteOnExit();
 
     ServerRecommender recommender = null;
@@ -132,10 +132,10 @@ private Multimap<Long,RecommendedItem> split(File dataDir,
 
     Multimap<Long,RecommendedItem> testData = ArrayListMultimap.create();
     Writer trainingOut =
-        new OutputStreamWriter(new BufferedOutputStream(new FileOutputStream(trainingFile)), Charsets.UTF_8);
+        new OutputStreamWriter(new GZIPOutputStream(new FileOutputStream(trainingFile)), Charsets.UTF_8);
     try {
 
-      Iterator<Map.Entry<Long, Collection<RecommendedItem>>> it = data.asMap().entrySet().iterator();
+      Iterator<Map.Entry<Long,Collection<RecommendedItem>>> it = data.asMap().entrySet().iterator();
       while (it.hasNext()) {
 
         Map.Entry<Long, Collection<RecommendedItem>> entry = it.next();
@@ -145,12 +145,7 @@ private Multimap<Long,RecommendedItem> split(File dataDir,
 
         if (isSplitTestByPrefValue()) {
           // Sort low to high, leaving high values at end for testing as "relevant" items
-          Collections.sort(userPrefs, new Comparator<RecommendedItem>() {
-            @Override
-            public int compare(RecommendedItem a, RecommendedItem b) {
-              return a.getValue() < b.getValue() ? -1 : a.getValue() > b.getValue() ? 1 : 0;
-            }
-          });
+          Collections.sort(userPrefs, ByValueAscComparator.INSTANCE);
         }
         // else leave sorted in time order