Package org.grouplens.lenskit.vectors

Examples of org.grouplens.lenskit.vectors.SparseVector

79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
        return new Context(outputLayout.prefixTable(tableWriter, algo, ds));
    }

    @Override
    public Void doMeasureUser(TestUser user, Context context) {
        SparseVector ratings = user.getTestRatings();
        SparseVector predictions = user.getPredictions();
        if (predictions == null) {
            predictions = MutableSparseVector.create();
        }

        LongSortedSet items = ratings.keySet();
        if (!items.containsAll(predictions.keySet())) {
            items = LongUtils.setUnion(items, predictions.keySet());
        }

        logger.debug("outputting {} predictions for user {}", predictions.size(), user.getUserId());
        for (LongIterator iter = items.iterator(); iter.hasNext(); /* no increment */) {
            long item = iter.nextLong();
            Double rating = ratings.containsKey(item) ? ratings.get(item) : null;
            Double pred = predictions.containsKey(item) ? predictions.get(item) : null;
            try {
                List<Object> row = Lists.<Object>newArrayList(user.getUserId(), item, rating, pred);
                for (Pair<Symbol, String> chan: channels) {
                    SparseVector v = predictions.getChannelVector(chan.getLeft());
                    if (v != null && v.containsKey(item)) {
                        row.add(v.get(item));
                    } else {
                        row.add(null);
                    }
                }
                context.writer.writeRow(row);
View Full Code Here
113
114
115
116
117
118
119
120
121
122
123
124
        recommendations = user.getRecommendations(listSize, candidates, exclude);
        if (recommendations == null) {
            return null;
        }

        SparseVector ratings = user.getTestRatings();
        LongList ideal = ratings.keysByValue(true);
        if (ideal.size() > listSize) {
            ideal = ideal.subList(0, listSize);
        }
        double idealGain = computeDCG(ideal, ratings);
View Full Code Here
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
    public SlopeOneModel get() {
        LongSet items = buildContext.getItems();
        LongIterator outer = items.iterator();
        while (outer.hasNext()) {
            final long item1 = outer.nextLong();
            final SparseVector vec1 = buildContext.itemVector(item1);
            LongIterator inner = items.iterator();
            while (inner.hasNext()) {
                final long item2 = inner.nextLong();
                if (item1 != item2) {
                    SparseVector vec2 = buildContext.itemVector(item2);
                    accumulator.putItemPair(item1, vec1, item2, vec2);
                }
            }
        }
        return new SlopeOneModel(accumulator.buildMatrix());
View Full Code Here
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
    public void score(long uid, @Nonnull MutableSparseVector scores) {
        UserHistory<Rating> history = dao.getEventsForUser(uid, Rating.class);
        if (history == null) {
            history = History.forUser(uid);
        }
        SparseVector user = RatingVectorUserHistorySummarizer.makeRatingVector(history);

        for (VectorEntry e : scores.view(VectorEntry.State.EITHER)) {
            final long predicteeItem = e.getKey();
            if (!user.containsKey(predicteeItem)) {
                double total = 0;
                int nitems = 0;
                LongIterator ratingIter = user.keySet().iterator();
                while (ratingIter.hasNext()) {
                    long currentItem = ratingIter.nextLong();
                    int nusers = model.getCoratings(predicteeItem, currentItem);
                    if (nusers != 0) {
                        double currentDev = model.getDeviation(predicteeItem, currentItem);
                        total += currentDev + user.get(currentItem);
                        nitems++;
                    }
                }
                if (nitems != 0) {
                    double predValue = total / nitems;
View Full Code Here
74
75
76
77
78
79
80
81
    }

    @Test
    public void testMeanBaseline() {
        ItemScorer pred = makeGlobalMean();
        SparseVector pv = pred.score(10L, itemSet(2l));
        assertEquals(RATINGS_DAT_MEAN, pv.get(2l), 0.00001);
    }
View Full Code Here
110
111
112
113
114
115
116
117
118
119
120
    @Test
    public void testItemMeanBaseline() {
        ItemScorer pred = new ItemMeanRatingItemScorer.Builder(dao, 0.0).get();
        long[] items = {5, 7, 10};
        double[] values = {3, 6, 4};
        SparseVector map = MutableSparseVector.wrap(items, values).freeze();
        // unseen item, should be global mean
        assertThat(pred.score(10, 2),
                   closeTo(RATINGS_DAT_MEAN, 0.001));
        // seen item - should be item average
        assertThat(pred.score(10, 5),
View Full Code Here
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
    public double getDeviation(long item1, long item2) {
        if (item1 == item2) {
            return 0;
        } else if (item1 < item2) {
            SparseVector row = matrix.get(item1);
            if (row == null) {
                return Double.NaN;
            } else {
                return row.get(item2);
            }
        } else {
            SparseVector row = matrix.get(item2);
            if (row == null) {
                return Double.NaN;
            } else {
                return -row.get(item1);
            }
        }
    }
View Full Code Here
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
    public int getCoratings(long item1, long item2) {
        if (item1 == item2) {
            return 0;
        } else if (item1 < item2) {
            SparseVector row = matrix.get(item1);
            if (row == null) {
                return 0;
            } else {
                double coratings = row.getChannelVector(CORATINGS_SYMBOL).get(item2, 0);
                return (int) coratings;
            }
        } else {
            SparseVector row = matrix.get(item2);
            if (row == null) {
                return 0;
            } else {
                double coratings = row.getChannelVector(CORATINGS_SYMBOL).get(item1, 0);
                return (int) coratings;
            }
        }
    }
View Full Code Here
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
    public void score(long uid, @Nonnull MutableSparseVector scores) {
        UserHistory<Rating> history = dao.getEventsForUser(uid, Rating.class);
        if (history == null) {
            history = History.forUser(uid);
        }
        SparseVector ratings = RatingVectorUserHistorySummarizer.makeRatingVector(history);

        for (VectorEntry e : scores.view(VectorEntry.State.EITHER)) {
            final long predicteeItem = e.getKey();
            if (!ratings.containsKey(predicteeItem)) {
                double total = 0;
                int nusers = 0;
                LongIterator ratingIter = ratings.keySet().iterator();
                while (ratingIter.hasNext()) {
                    long currentItem = ratingIter.nextLong();
                    double currentDev = model.getDeviation(predicteeItem, currentItem);
                    if (!Double.isNaN(currentDev)) {
                        int weight = model.getCoratings(predicteeItem, currentItem);
                        total += (currentDev + ratings.get(currentItem)) * weight;
                        nusers += weight;
                    }
                }
                if (nusers == 0) {
                    scores.unset(e);
View Full Code Here
46
47
48
49
50
51
52
53
54
    }

    @Test
    public void testScoreSet() {
        ItemScorer pred = new ConstantItemScorer(5);
        SparseVector v = pred.score(42, LongUtils.packedSet(1, 2, 3, 5, 7));
        assertThat(v.keySet(), contains(1L, 2L, 3L, 5L, 7L));
        assertThat(v.values(), everyItem(equalTo(5.0)));
    }
View Full Code Here
TOP

Related Classes of org.grouplens.lenskit.vectors.SparseVector

  • it.unimi.dsi.fastutil.doubles.DoubleArrayList
  • it.unimi.dsi.fastutil.doubles.DoubleIterator
  • it.unimi.dsi.fastutil.ints.IntIterator
  • org.grouplens.lenskit.baseline.ConstantItemScorerTest
  • org.grouplens.lenskit.baseline.MeanScorersTest
  • org.grouplens.lenskit.basic.AbstractGlobalItemScorer
  • org.grouplens.lenskit.basic.AbstractItemScorer
  • org.grouplens.lenskit.basic.AbstractRatingPredictor
  • org.grouplens.lenskit.basic.PrecomputedItemScorer
  • org.grouplens.lenskit.basic.RescoringItemRecommender
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.