Package hivemall.io

Examples of hivemall.io.PredictionModel

145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
    protected PredictionModel createModel() {
        return createModel(null);
    }

    protected PredictionModel createModel(String label) {
        PredictionModel model;
        final boolean useCovar = useCovariance();
        if(dense_model) {
            if(disable_halffloat == false && model_dims > 16777216) {
                logger.info("Build a space efficient dense model with " + model_dims
                        + " initial dimensions" + (useCovar ? " w/ covariances" : ""));
                model = new SpaceEfficientDenseModel(model_dims, useCovar);
            } else {
                logger.info("Build a dense model with initial with " + model_dims
                        + " initial dimensions" + (useCovar ? " w/ covariances" : ""));
                model = new DenseModel(model_dims, useCovar);
            }
        } else {
            int initModelSize = getInitialModelSize();
            logger.info("Build a sparse model with initial with " + initModelSize
                    + " initial dimensions");
            model = new SparseModel(initModelSize, useCovar);
        }
        if(mixConnectInfo != null) {
            model.configureClock();
            model = new SynchronizedModelWrapper(model);
            MixClient client = configureMixClient(mixConnectInfo, label, model);
            model.setUpdateHandler(client);
            this.mixClient = client;
        }
        assert (model != null);
        return model;
    }
View Full Code Here
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
        ExecutorService serverExec = Executors.newSingleThreadExecutor();
        serverExec.submit(server);

        waitForState(server, ServerState.RUNNING);

        PredictionModel model = new DenseModel(16777216, false);
        model.configureClock();
        MixClient client = null;
        try {
            client = new MixClient(MixEventName.average, "testSimpleScenario", "localhost:" + port, false, 2, model);
            model.setUpdateHandler(client);

            final Random rand = new Random(43);
            for(int i = 0; i < 100000; i++) {
                Integer feature = Integer.valueOf(rand.nextInt(100));
                float weight = (float) rand.nextGaussian();
                model.set(feature, new WeightValue(weight));
            }

            waitForMixed(model, 48000, 10000L);

            int numMixed = model.getNumMixed();
            //System.out.println("number of mix events: " + numMixed);
            Assert.assertTrue("number of mix events: " + numMixed, numMixed > 0);

            serverExec.shutdown();
        } finally {
View Full Code Here
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
        ExecutorService serverExec = Executors.newSingleThreadExecutor();
        serverExec.submit(server);

        waitForState(server, ServerState.RUNNING);

        PredictionModel model = new DenseModel(16777216, false);
        model.configureClock();
        MixClient client = null;
        try {
            client = new MixClient(MixEventName.average, "testSSL", "localhost:" + port, true, 2, model);
            model.setUpdateHandler(client);

            final Random rand = new Random(43);
            for(int i = 0; i < 100000; i++) {
                Integer feature = Integer.valueOf(rand.nextInt(100));
                float weight = (float) rand.nextGaussian();
                model.set(feature, new WeightValue(weight));
            }

            waitForMixed(model, 48000, 10000L);

            int numMixed = model.getNumMixed();
            //System.out.println("number of mix events: " + numMixed);
            Assert.assertTrue("number of mix events: " + numMixed, numMixed > 0);

            serverExec.shutdown();
        } finally {
View Full Code Here
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
        clientsExec.shutdown();
        serverExec.shutdown();
    }

    private static void invokeClient(String groupId, int serverPort) throws InterruptedException {
        PredictionModel model = new DenseModel(16777216, false);
        model.configureClock();
        MixClient client = null;
        try {
            client = new MixClient(MixEventName.average, groupId, "localhost:" + serverPort, false, 2, model);
            model.setUpdateHandler(client);

            final Random rand = new Random(43);
            for(int i = 0; i < 100000; i++) {
                Integer feature = Integer.valueOf(rand.nextInt(100));
                float weight = (float) rand.nextGaussian();
                model.set(feature, new WeightValue(weight));
            }

            waitForMixed(model, 48000, 10000L);

            int numMixed = model.getNumMixed();
            //System.out.println("number of mix events: " + numMixed);
            Assert.assertTrue("number of mix events: " + numMixed, numMixed > 0);
        } finally {
            IOUtils.closeQuietly(client);
        }
View Full Code Here
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
        serverExec.shutdown();
    }

    private static void invokeClient01(String groupId, int serverPort, boolean denseModel)
            throws InterruptedException {
        PredictionModel model = denseModel ? new DenseModel(100, false)
                : new SparseModel(100, false);
        model.configureClock();
        MixClient client = null;
        try {
            client = new MixClient(MixEventName.average, groupId, "localhost:" + serverPort, false, 3, model);
            model.setUpdateHandler(client);

            final Random rand = new Random(43);
            for(int i = 0; i < 1000000; i++) {
                Integer feature = Integer.valueOf(rand.nextInt(100));
                float weight = rand.nextFloat() >= 0.5f ? 1.f : 0.f;
                model.set(feature, new WeightValue(weight));
            }

            waitForMixed(model, 100000, 10000L);

            int numMixed = model.getNumMixed();
            //System.out.println("number of mix events: " + numMixed);
            Assert.assertTrue("number of mix events: " + numMixed, numMixed > 0);

            for(int i = 0; i < 100; i++) {
                float w = model.getWeight(i);
                Assert.assertEquals(0.5f, w, 0.1f);
            }
        } finally {
            IOUtils.closeQuietly(client);
        }
View Full Code Here
TOP

Related Classes of hivemall.io.PredictionModel

  • hivemall.classifier.multiclass.MulticlassAROWClassifierUDTF
  • hivemall.classifier.multiclass.MulticlassConfidenceWeightedUDTF
  • hivemall.classifier.multiclass.MulticlassOnlineClassifierUDTF
  • hivemall.classifier.multiclass.MulticlassSoftConfidenceWeightedUDTF
  • hivemall.LearnerBaseUDTF
  • hivemall.mix.server.MixServerTest
    • 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.