List of usage examples for org.apache.mahout.math Vector times
Vector times(Vector x);
From source file:DisplayClustering.java
License:Apache License
/** * Identical to plotRectangle(), but with the option of setting the color of * the rectangle's stroke.// w ww. j a v a 2 s. c o m * * NOTE: This should probably be refactored with plotRectangle() since most of * the code here is direct copy/paste from that method. * * @param g2 * A Graphics2D context. * @param v * A vector for the rectangle's center. * @param dv * A vector for the rectangle's dimensions. * @param color * The color of the rectangle's stroke. */ protected static void plotClusteredRectangle(Graphics2D g2, Vector v, Vector dv, Color color) { double[] flip = { 1, -1 }; Vector v2 = v.times(new DenseVector(flip)); v2 = v2.minus(dv.divide(2)); int h = SIZE / 2; double x = v2.get(0) + h; double y = v2.get(1) + h; g2.setStroke(new BasicStroke(1)); g2.setColor(color); g2.draw(new Rectangle2D.Double(x * DS, y * DS, dv.get(0) * DS, dv.get(1) * DS)); }
From source file:DisplayClustering.java
License:Apache License
/** * Draw a rectangle on the graphics context * * @param g2/*from w ww . jav a 2 s. c o m*/ * a Graphics2D context * @param v * a Vector of rectangle center * @param dv * a Vector of rectangle dimensions */ protected static void plotRectangle(Graphics2D g2, Vector v, Vector dv) { double[] flip = { 1, -1 }; Vector v2 = v.times(new DenseVector(flip)); v2 = v2.minus(dv.divide(2)); int h = SIZE / 2; double x = v2.get(0) + h; double y = v2.get(1) + h; g2.draw(new Rectangle2D.Double(x * DS, y * DS, dv.get(0) * DS, dv.get(1) * DS)); }
From source file:DisplayClustering.java
License:Apache License
/** * Draw an ellipse on the graphics context * * @param g2/* w w w . j a v a 2s . co m*/ * a Graphics2D context * @param v * a Vector of ellipse center * @param dv * a Vector of ellipse dimensions */ protected static void plotEllipse(Graphics2D g2, Vector v, Vector dv) { double[] flip = { 1, -1 }; Vector v2 = v.times(new DenseVector(flip)); v2 = v2.minus(dv.divide(2)); int h = SIZE / 2; double x = v2.get(0) + h; double y = v2.get(1) + h; g2.draw(new Ellipse2D.Double(x * DS, y * DS, dv.get(0) * DS, dv.get(1) * DS)); }
From source file:at.illecker.hama.rootbeer.examples.matrixmultiplication.compositeinput.cpu.MatrixMultiplicationBSPCpu.java
License:Apache License
@Override public void bsp(BSPPeer<IntWritable, TupleWritable, IntWritable, VectorWritable, MatrixRowMessage> peer) throws IOException, SyncException, InterruptedException { IntWritable key = new IntWritable(); TupleWritable value = new TupleWritable(); while (peer.readNext(key, value)) { // Logging if (isDebuggingEnabled) { for (int i = 0; i < value.size(); i++) { Vector vector = ((VectorWritable) value.get(i)).get(); logger.writeChars("bsp,input,key=" + key + ",value=" + vector.toString() + "\n"); }//from www.j a v a2 s . c o m } Vector firstVector = ((VectorWritable) value.get(0)).get(); Vector secondVector = ((VectorWritable) value.get(1)).get(); // outCardinality is resulting column size n // (l x m) * (m x n) = (l x n) boolean firstIsOutFrag = secondVector.size() == outCardinality; // outFrag is Matrix which has the resulting column cardinality // (matrixB) Vector outFrag = firstIsOutFrag ? secondVector : firstVector; // multiplier is Matrix which has the resulting row count // (transposed matrixA) Vector multiplier = firstIsOutFrag ? firstVector : secondVector; if (isDebuggingEnabled) { logger.writeChars("bsp,firstIsOutFrag=" + firstIsOutFrag + "\n"); logger.writeChars("bsp,outFrag=" + outFrag + "\n"); logger.writeChars("bsp,multiplier=" + multiplier + "\n"); } for (Vector.Element e : multiplier.nonZeroes()) { VectorWritable outVector = new VectorWritable(); // Scalar Multiplication (Vector x Element) outVector.set(outFrag.times(e.get())); peer.send(masterTask, new MatrixRowMessage(e.index(), outVector)); if (isDebuggingEnabled) { logger.writeChars("bsp,send,key=" + e.index() + ",value=" + outVector.get().toString() + "\n"); } } if (isDebuggingEnabled) { logger.flush(); } } peer.sync(); }
From source file:ca.uwaterloo.cpami.mahout.matrix.utils.GramSchmidt.java
License:Apache License
public static void orthonormalizeColumns(Matrix mx) { //int n = mx.numCols(); int n = mx.numRows(); for (int c = 0; c < n; c++) { System.out.println("col: " + c); Vector col = mx.viewRow(c); for (int c1 = 0; c1 < c; c1++) { Vector viewC1 = mx.viewRow(c1); col.assign(col.minus(viewC1.times(viewC1.dot(col)))); }//from w w w. jav a2 s. c o m final double norm2 = col.norm(2); if (norm2 == 0) { System.out.println("zero"); } col.assign(new DoubleFunction() { @Override public double apply(double x) { return x / norm2; } }); } }
From source file:cn.edu.bjtu.cit.recommender.Recommender.java
License:Apache License
@SuppressWarnings("unchecked") public int run(String[] args) throws Exception { if (args.length < 2) { System.err.println();//from w ww. ja v a 2 s . com System.err.println("Usage: " + this.getClass().getName() + " [generic options] input output [profiling] [estimation] [clustersize]"); System.err.println(); printUsage(); GenericOptionsParser.printGenericCommandUsage(System.err); return 1; } OptionParser parser = new OptionParser(args); Pipeline pipeline = new MRPipeline(Recommender.class, getConf()); if (parser.hasOption(CLUSTER_SIZE)) { pipeline.getConfiguration().setInt(ClusterOracle.CLUSTER_SIZE, Integer.parseInt(parser.getOption(CLUSTER_SIZE).getValue())); } if (parser.hasOption(PROFILING)) { pipeline.getConfiguration().setBoolean(Profiler.IS_PROFILE, true); this.profileFilePath = parser.getOption(PROFILING).getValue(); } if (parser.hasOption(ESTIMATION)) { estFile = parser.getOption(ESTIMATION).getValue(); est = new Estimator(estFile, clusterSize); } if (parser.hasOption(OPT_REDUCE)) { pipeline.getConfiguration().setBoolean(OPT_REDUCE, true); } if (parser.hasOption(OPT_MSCR)) { pipeline.getConfiguration().setBoolean(OPT_MSCR, true); } if (parser.hasOption(ACTIVE_THRESHOLD)) { threshold = Integer.parseInt(parser.getOption("at").getValue()); } if (parser.hasOption(TOP)) { top = Integer.parseInt(parser.getOption("top").getValue()); } profiler = new Profiler(pipeline); /* * input node */ PCollection<String> lines = pipeline.readTextFile(args[0]); if (profiler.isProfiling() && lines.getSize() > 10 * 1024 * 1024) { lines = lines.sample(0.1); } /* * S0 + GBK */ PGroupedTable<Long, Long> userWithPrefs = lines.parallelDo(new MapFn<String, Pair<Long, Long>>() { @Override public Pair<Long, Long> map(String input) { String[] split = input.split(Estimator.DELM); long userID = Long.parseLong(split[0]); long itemID = Long.parseLong(split[1]); return Pair.of(userID, itemID); } @Override public float scaleFactor() { return est.getScaleFactor("S0").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("S0").recsFactor; } }, Writables.tableOf(Writables.longs(), Writables.longs())).groupByKey(est.getClusterSize()); /* * S1 */ PTable<Long, Vector> userVector = userWithPrefs .parallelDo(new MapFn<Pair<Long, Iterable<Long>>, Pair<Long, Vector>>() { @Override public Pair<Long, Vector> map(Pair<Long, Iterable<Long>> input) { Vector userVector = new RandomAccessSparseVector(Integer.MAX_VALUE, 100); for (long itemPref : input.second()) { userVector.set((int) itemPref, 1.0f); } return Pair.of(input.first(), userVector); } @Override public float scaleFactor() { return est.getScaleFactor("S1").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("S1").recsFactor; } }, Writables.tableOf(Writables.longs(), Writables.vectors())); userVector = profiler.profile("S0-S1", pipeline, userVector, ProfileConverter.long_vector(), Writables.tableOf(Writables.longs(), Writables.vectors())); /* * S2 */ PTable<Long, Vector> filteredUserVector = userVector .parallelDo(new DoFn<Pair<Long, Vector>, Pair<Long, Vector>>() { @Override public void process(Pair<Long, Vector> input, Emitter<Pair<Long, Vector>> emitter) { if (input.second().getNumNondefaultElements() > threshold) { emitter.emit(input); } } @Override public float scaleFactor() { return est.getScaleFactor("S2").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("S2").recsFactor; } }, Writables.tableOf(Writables.longs(), Writables.vectors())); filteredUserVector = profiler.profile("S2", pipeline, filteredUserVector, ProfileConverter.long_vector(), Writables.tableOf(Writables.longs(), Writables.vectors())); /* * S3 + GBK */ PGroupedTable<Integer, Integer> coOccurencePairs = filteredUserVector .parallelDo(new DoFn<Pair<Long, Vector>, Pair<Integer, Integer>>() { @Override public void process(Pair<Long, Vector> input, Emitter<Pair<Integer, Integer>> emitter) { Iterator<Vector.Element> it = input.second().iterateNonZero(); while (it.hasNext()) { int index1 = it.next().index(); Iterator<Vector.Element> it2 = input.second().iterateNonZero(); while (it2.hasNext()) { int index2 = it2.next().index(); emitter.emit(Pair.of(index1, index2)); } } } @Override public float scaleFactor() { float size = est.getScaleFactor("S3").sizeFactor; return size; } @Override public float scaleFactorByRecord() { float recs = est.getScaleFactor("S3").recsFactor; return recs; } }, Writables.tableOf(Writables.ints(), Writables.ints())).groupByKey(est.getClusterSize()); /* * S4 */ PTable<Integer, Vector> coOccurenceVector = coOccurencePairs .parallelDo(new MapFn<Pair<Integer, Iterable<Integer>>, Pair<Integer, Vector>>() { @Override public Pair<Integer, Vector> map(Pair<Integer, Iterable<Integer>> input) { Vector cooccurrenceRow = new RandomAccessSparseVector(Integer.MAX_VALUE, 100); for (int itemIndex2 : input.second()) { cooccurrenceRow.set(itemIndex2, cooccurrenceRow.get(itemIndex2) + 1.0); } return Pair.of(input.first(), cooccurrenceRow); } @Override public float scaleFactor() { return est.getScaleFactor("S4").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("S4").recsFactor; } }, Writables.tableOf(Writables.ints(), Writables.vectors())); coOccurenceVector = profiler.profile("S3-S4", pipeline, coOccurenceVector, ProfileConverter.int_vector(), Writables.tableOf(Writables.ints(), Writables.vectors())); /* * S5 Wrapping co-occurrence columns */ PTable<Integer, VectorOrPref> wrappedCooccurrence = coOccurenceVector .parallelDo(new MapFn<Pair<Integer, Vector>, Pair<Integer, VectorOrPref>>() { @Override public Pair<Integer, VectorOrPref> map(Pair<Integer, Vector> input) { return Pair.of(input.first(), new VectorOrPref(input.second())); } @Override public float scaleFactor() { return est.getScaleFactor("S5").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("S5").recsFactor; } }, Writables.tableOf(Writables.ints(), VectorOrPref.vectorOrPrefs())); wrappedCooccurrence = profiler.profile("S5", pipeline, wrappedCooccurrence, ProfileConverter.int_vopv(), Writables.tableOf(Writables.ints(), VectorOrPref.vectorOrPrefs())); /* * S6 Splitting user vectors */ PTable<Integer, VectorOrPref> userVectorSplit = filteredUserVector .parallelDo(new DoFn<Pair<Long, Vector>, Pair<Integer, VectorOrPref>>() { @Override public void process(Pair<Long, Vector> input, Emitter<Pair<Integer, VectorOrPref>> emitter) { long userID = input.first(); Vector userVector = input.second(); Iterator<Vector.Element> it = userVector.iterateNonZero(); while (it.hasNext()) { Vector.Element e = it.next(); int itemIndex = e.index(); float preferenceValue = (float) e.get(); emitter.emit(Pair.of(itemIndex, new VectorOrPref(userID, preferenceValue))); } } @Override public float scaleFactor() { return est.getScaleFactor("S6").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("S6").recsFactor; } }, Writables.tableOf(Writables.ints(), VectorOrPref.vectorOrPrefs())); userVectorSplit = profiler.profile("S6", pipeline, userVectorSplit, ProfileConverter.int_vopp(), Writables.tableOf(Writables.ints(), VectorOrPref.vectorOrPrefs())); /* * S7 Combine VectorOrPrefs */ PTable<Integer, VectorAndPrefs> combinedVectorOrPref = wrappedCooccurrence.union(userVectorSplit) .groupByKey(est.getClusterSize()) .parallelDo(new DoFn<Pair<Integer, Iterable<VectorOrPref>>, Pair<Integer, VectorAndPrefs>>() { @Override public void process(Pair<Integer, Iterable<VectorOrPref>> input, Emitter<Pair<Integer, VectorAndPrefs>> emitter) { Vector vector = null; List<Long> userIDs = Lists.newArrayList(); List<Float> values = Lists.newArrayList(); for (VectorOrPref vop : input.second()) { if (vector == null) { vector = vop.getVector(); } long userID = vop.getUserID(); if (userID != Long.MIN_VALUE) { userIDs.add(vop.getUserID()); } float value = vop.getValue(); if (!Float.isNaN(value)) { values.add(vop.getValue()); } } emitter.emit(Pair.of(input.first(), new VectorAndPrefs(vector, userIDs, values))); } @Override public float scaleFactor() { return est.getScaleFactor("S7").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("S7").recsFactor; } }, Writables.tableOf(Writables.ints(), VectorAndPrefs.vectorAndPrefs())); combinedVectorOrPref = profiler.profile("S5+S6-S7", pipeline, combinedVectorOrPref, ProfileConverter.int_vap(), Writables.tableOf(Writables.ints(), VectorAndPrefs.vectorAndPrefs())); /* * S8 Computing partial recommendation vectors */ PTable<Long, Vector> partialMultiply = combinedVectorOrPref .parallelDo(new DoFn<Pair<Integer, VectorAndPrefs>, Pair<Long, Vector>>() { @Override public void process(Pair<Integer, VectorAndPrefs> input, Emitter<Pair<Long, Vector>> emitter) { Vector cooccurrenceColumn = input.second().getVector(); List<Long> userIDs = input.second().getUserIDs(); List<Float> prefValues = input.second().getValues(); for (int i = 0; i < userIDs.size(); i++) { long userID = userIDs.get(i); if (userID != Long.MIN_VALUE) { float prefValue = prefValues.get(i); Vector partialProduct = cooccurrenceColumn.times(prefValue); emitter.emit(Pair.of(userID, partialProduct)); } } } @Override public float scaleFactor() { return est.getScaleFactor("S8").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("S8").recsFactor; } }, Writables.tableOf(Writables.longs(), Writables.vectors())).groupByKey(est.getClusterSize()) .combineValues(new CombineFn<Long, Vector>() { @Override public void process(Pair<Long, Iterable<Vector>> input, Emitter<Pair<Long, Vector>> emitter) { Vector partial = null; for (Vector vector : input.second()) { partial = partial == null ? vector : partial.plus(vector); } emitter.emit(Pair.of(input.first(), partial)); } @Override public float scaleFactor() { return est.getScaleFactor("combine").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("combine").recsFactor; } }); partialMultiply = profiler.profile("S8-combine", pipeline, partialMultiply, ProfileConverter.long_vector(), Writables.tableOf(Writables.longs(), Writables.vectors())); /* * S9 Producing recommendations from vectors */ PTable<Long, RecommendedItems> recommendedItems = partialMultiply .parallelDo(new DoFn<Pair<Long, Vector>, Pair<Long, RecommendedItems>>() { @Override public void process(Pair<Long, Vector> input, Emitter<Pair<Long, RecommendedItems>> emitter) { Queue<RecommendedItem> topItems = new PriorityQueue<RecommendedItem>(11, Collections.reverseOrder(BY_PREFERENCE_VALUE)); Iterator<Vector.Element> recommendationVectorIterator = input.second().iterateNonZero(); while (recommendationVectorIterator.hasNext()) { Vector.Element element = recommendationVectorIterator.next(); int index = element.index(); float value = (float) element.get(); if (topItems.size() < top) { topItems.add(new GenericRecommendedItem(index, value)); } else if (value > topItems.peek().getValue()) { topItems.add(new GenericRecommendedItem(index, value)); topItems.poll(); } } List<RecommendedItem> recommendations = new ArrayList<RecommendedItem>(topItems.size()); recommendations.addAll(topItems); Collections.sort(recommendations, BY_PREFERENCE_VALUE); emitter.emit(Pair.of(input.first(), new RecommendedItems(recommendations))); } @Override public float scaleFactor() { return est.getScaleFactor("S9").sizeFactor; } @Override public float scaleFactorByRecord() { return est.getScaleFactor("S9").recsFactor; } }, Writables.tableOf(Writables.longs(), RecommendedItems.recommendedItems())); recommendedItems = profiler.profile("S9", pipeline, recommendedItems, ProfileConverter.long_ri(), Writables.tableOf(Writables.longs(), RecommendedItems.recommendedItems())); /* * Profiling */ if (profiler.isProfiling()) { profiler.writeResultToFile(profileFilePath); profiler.cleanup(pipeline.getConfiguration()); return 0; } /* * asText */ pipeline.writeTextFile(recommendedItems, args[1]); PipelineResult result = pipeline.done(); return result.succeeded() ? 0 : 1; }
From source file:com.cloudera.science.ml.kmeans.core.MiniBatchUpdateStrategy.java
License:Open Source License
@Override public <V extends Vector> Centers update(List<Weighted<V>> points, Centers centers) { int[] perCenterStepCounts = new int[centers.size()]; WeightedSampler<V> sampler = new WeightedSampler<V>(points, random); for (int iter = 0; iter < numIterations; iter++) { // Compute closest cent for each mini-batch List<List<V>> centerAssignments = Lists.newArrayList(); for (int i = 0; i < centers.size(); i++) { centerAssignments.add(Lists.<V>newArrayList()); }/* ww w .j a v a2s . c o m*/ for (int i = 0; i < miniBatchSize; i++) { V sample = sampler.sample(); int closestId = centers.indexOfClosest(sample); centerAssignments.get(closestId).add(sample); } // Apply the mini-batch List<Vector> nextCenters = Lists.newArrayList(); for (int i = 0; i < centerAssignments.size(); i++) { Vector currentCenter = centers.get(i); for (int j = 0; j < centerAssignments.get(i).size(); j++) { double eta = 1.0 / (++perCenterStepCounts[i] + 1.0); currentCenter = currentCenter.times(1.0 - eta); currentCenter = currentCenter.plus(centerAssignments.get(i).get(j).times(eta)); } nextCenters.add(currentCenter); } centers = new Centers(nextCenters); } return centers; }
From source file:com.ydy.cf.solver.impl.AlternatingLeastSquaresImplicitSolver.java
License:Apache License
/** Y' (Cu - I) Y + I */ private Matrix YtransponseCuMinusIYPlusLambdaI(Vector userRatings) { Preconditions.checkArgument(userRatings.isSequentialAccess(), "need sequential access to ratings!"); /* (Cu -I) Y */ OpenIntObjectHashMap<Vector> CuMinusIY = new OpenIntObjectHashMap<Vector>(); Iterator<Vector.Element> ratings = userRatings.iterateNonZero(); while (ratings.hasNext()) { Vector.Element e = ratings.next(); Vector curYRow = Y.viewRow(e.index()); CuMinusIY.put(e.index(), curYRow.times(confidence(e.get()) - 1)); }/* w w w . jav a 2 s .c o m*/ Matrix YtransponseCuMinusIY = new DenseMatrix(numFeatures, numFeatures); /* Y' (Cu -I) Y by outer products */ ratings = userRatings.iterateNonZero(); while (ratings.hasNext()) { Vector.Element e = ratings.next(); for (Vector.Element feature : Y.viewRow(e.index())) { Vector partial = CuMinusIY.get(e.index()).times(feature.get()); YtransponseCuMinusIY.viewRow(feature.index()).assign(partial, Functions.PLUS); } } /* Y' (Cu - I) Y + I add lambda on the diagonal */ for (int feature = 0; feature < numFeatures; feature++) { YtransponseCuMinusIY.setQuick(feature, feature, YtransponseCuMinusIY.getQuick(feature, feature) + lambda); } return YtransponseCuMinusIY; }
From source file:com.ydy.cf.solver.impl.AlternatingLeastSquaresImplicitSolver.java
License:Apache License
/** Y' Cu p(u) */ private Matrix YtransponseCuPu(Vector userRatings) { Preconditions.checkArgument(userRatings.isSequentialAccess(), "need sequential access to ratings!"); Vector YtransponseCuPu = new DenseVector(numFeatures); Iterator<Vector.Element> ratings = userRatings.iterateNonZero(); while (ratings.hasNext()) { Vector.Element e = ratings.next(); Vector curYRow = Y.viewRow(e.index()); YtransponseCuPu.assign(curYRow.times(confidence(e.get())), Functions.PLUS); }/*from ww w .j a v a 2s .c om*/ return columnVectorAsMatrix(YtransponseCuPu); }
From source file:edu.indiana.d2i.htrc.kmeans.KMeansClusterer.java
License:Apache License
/** * Iterates over all clusters and identifies the one closes to the given point. Distance measure used is * configured at creation time./*from w w w . j av a2 s .c o m*/ * * @param point * a point to find a cluster for. * @param clusters * a List<Cluster> to test. */ public void emitPointToNearestCluster(Vector point, Iterable<Cluster> clusters, Mapper<?, ?, Text, ClusterObservations>.Context context) throws IOException, InterruptedException { Cluster nearestCluster = null; double nearestDistance = Double.MAX_VALUE; for (Cluster cluster : clusters) { Vector clusterCenter = cluster.getCenter(); double distance = this.measure.distance(clusterCenter.getLengthSquared(), clusterCenter, point); if (log.isDebugEnabled()) { log.debug("{} Cluster: {}", distance, cluster.getId()); } if (distance < nearestDistance || nearestCluster == null) { nearestCluster = cluster; nearestDistance = distance; } } context.write(new Text(nearestCluster.getIdentifier()), new ClusterObservations(1, point, point.times(point))); }