Examples with TwoDimTable water.util.TwoDimTable used on opensource projects

Example 26 with TwoDimTable

use of water.util.TwoDimTable in project h2o-3 by h2oai.

the class ModelMetricsMultinomial method getHitRatioTable.

public static TwoDimTable getHitRatioTable(float[] hits) {
    String tableHeader = "Top-" + hits.length + " Hit Ratios";
    String[] rowHeaders = new String[hits.length];
    for (int k = 0; k < hits.length; ++k) rowHeaders[k] = Integer.toString(k + 1);
    String[] colHeaders = new String[] { "Hit Ratio" };
    String[] colTypes = new String[] { "float" };
    String[] colFormats = new String[] { "%f" };
    String colHeaderForRowHeaders = "K";
    TwoDimTable table = new TwoDimTable(tableHeader, null, /*tableDescription*/
    rowHeaders, colHeaders, colTypes, colFormats, colHeaderForRowHeaders);
    for (int k = 0; k < hits.length; ++k) table.set(k, 0, hits[k]);
    return table;
}

Example 27 with TwoDimTable

use of water.util.TwoDimTable in project h2o-3 by h2oai.

the class GainsLift method createTwoDimTable.

public TwoDimTable createTwoDimTable() {
    if (response_rates == null || Double.isNaN(avg_response_rate))
        return null;
    TwoDimTable table = new TwoDimTable("Gains/Lift Table", "Avg response rate: " + PrettyPrint.formatPct(avg_response_rate), new String[events.length], new String[] { "Group", "Cumulative Data Fraction", "Lower Threshold", "Lift", "Cumulative Lift", "Response Rate", "Cumulative Response Rate", "Capture Rate", "Cumulative Capture Rate", "Gain", "Cumulative Gain" }, new String[] { "int", "double", "double", "double", "double", "double", "double", "double", "double", "double", "double" }, new String[] { "%d", "%.8f", "%5f", "%5f", "%5f", "%5f", "%5f", "%5f", "%5f", "%5f", "%5f" }, "");
    long sum_e_i = 0;
    long sum_n_i = 0;
    // E/N
    double P = avg_response_rate;
    long N = ArrayUtils.sum(observations);
    long E = Math.round(N * P);
    for (int i = 0; i < events.length; ++i) {
        long e_i = events[i];
        long n_i = observations[i];
        double p_i = response_rates[i];
        sum_e_i += e_i;
        sum_n_i += n_i;
        //can be NaN if P==0
        double lift = p_i / P;
        //can be NaN if P==0
        double sum_lift = (double) sum_e_i / sum_n_i / P;
        //group
        table.set(i, 0, i + 1);
        //cumulative_data_fraction
        table.set(i, 1, (double) sum_n_i / N);
        //lower_threshold
        table.set(i, 2, _quantiles[i]);
        //lift
        table.set(i, 3, lift);
        //cumulative_lift
        table.set(i, 4, sum_lift);
        //response_rate
        table.set(i, 5, p_i);
        //cumulative_response_rate
        table.set(i, 6, (double) sum_e_i / sum_n_i);
        //capture_rate
        table.set(i, 7, (double) e_i / E);
        //cumulative_capture_rate
        table.set(i, 8, (double) sum_e_i / E);
        //gain
        table.set(i, 9, 100 * (lift - 1));
        //cumulative gain
        table.set(i, 10, 100 * (sum_lift - 1));
        if (i == events.length - 1) {
            assert (sum_n_i == N) : "Cumulative data fraction must be 1.0, but is " + (double) sum_n_i / N;
            assert (sum_e_i == E) : "Cumulative capture rate must be 1.0, but is " + (double) sum_e_i / E;
            if (!Double.isNaN(sum_lift))
                assert (Math.abs(sum_lift - 1.0) < 1e-8) : "Cumulative lift must be 1.0, but is " + sum_lift;
            assert (Math.abs((double) sum_e_i / sum_n_i - avg_response_rate) < 1e-8) : "Cumulative response rate must be " + avg_response_rate + ", but is " + (double) sum_e_i / sum_n_i;
        }
    }
    return this.table = table;
}

Example 28 with TwoDimTable

use of water.util.TwoDimTable in project h2o-3 by h2oai.

the class DimensionReductionUtils method createScoringHistoryTableDR.

/**
     * Create the scoring history for dimension reduction algorithms like PCA/SVD.  We do make the following assumptions
     * about your scoring_history.  First we assume that you will always have the following field:
     * 1. Timestamp: long denoting the time in ms;
     * 2. All other fields are double.
     *
     * The following field will be generated for you automatically: Duration and Iteration.
     *
     * @param scoreTable: HashMap containing column headers and arraylist containing the history of values collected.
     * @param tableName: title/name of your scoring table
     * @param startTime:  time your model building job was first started.
     * @return: TwoDimTable containing the scoring history.
     */
public static TwoDimTable createScoringHistoryTableDR(LinkedHashMap<String, ArrayList> scoreTable, String tableName, long startTime) {
    List<String> colHeaders = new ArrayList<>();
    List<String> colTypes = new ArrayList<>();
    List<String> colFormat = new ArrayList<>();
    ArrayList<String> otherTableEntries = new ArrayList<String>();
    for (String fieldName : scoreTable.keySet()) {
        if (fieldName.equals("Timestamp")) {
            colHeaders.add("Timestamp");
            colTypes.add("string");
            colFormat.add("%s");
            colHeaders.add("Duration");
            colTypes.add("string");
            colFormat.add("%s");
            colHeaders.add("Iteration");
            colTypes.add("long");
            colFormat.add("%d");
        } else {
            otherTableEntries.add(fieldName);
            colHeaders.add(fieldName);
            colTypes.add("double");
            colFormat.add("%.5f");
        }
    }
    // number of entries of training history
    int rows = scoreTable.get("Timestamp").size();
    TwoDimTable table = new TwoDimTable(tableName, null, new String[rows], colHeaders.toArray(new String[0]), colTypes.toArray(new String[0]), colFormat.toArray(new String[0]), "");
    assert (rows <= table.getRowDim());
    for (int row = 0; row < rows; row++) {
        int col = 0;
        // take care of Timestamp, Duration, Iteration.
        DateTimeFormatter fmt = DateTimeFormat.forPattern("yyyy-MM-dd HH:mm:ss");
        table.set(row, col++, fmt.print((long) scoreTable.get("Timestamp").get(row)));
        table.set(row, col++, PrettyPrint.msecs((long) scoreTable.get("Timestamp").get(row) - startTime, true));
        table.set(row, col++, row);
        // take care of the extra field
        for (int remaining_cols = 0; remaining_cols < otherTableEntries.size(); remaining_cols++) {
            table.set(row, col++, (double) scoreTable.get(otherTableEntries.get(remaining_cols)).get(row));
        }
    }
    return table;
}

Example 29 with TwoDimTable

use of water.util.TwoDimTable in project h2o-3 by h2oai.

the class PartialDependenceTest method weatherBinary.

@Test
public void weatherBinary() {
    Frame fr = null;
    GBMModel model = null;
    PartialDependence partialDependence = null;
    try {
        // Frame
        fr = parse_test_file("smalldata/junit/weather.csv");
        // Model
        GBMModel.GBMParameters parms = new GBMModel.GBMParameters();
        parms._train = fr._key;
        parms._ignored_columns = new String[] { "Date", "RISK_MM", "EvapMM" };
        parms._response_column = "RainTomorrow";
        model = new GBM(parms).trainModel().get();
        // PartialDependence
        partialDependence = new PartialDependence(Key.<PartialDependence>make());
        partialDependence._nbins = 33;
        partialDependence._cols = new String[] { "Sunshine", "MaxWindPeriod", "WindSpeed9am" };
        partialDependence._model_id = (Key) model._key;
        partialDependence._frame_id = fr._key;
        partialDependence.execImpl().get();
        for (TwoDimTable t : partialDependence._partial_dependence_data) Log.info(t);
    } finally {
        if (fr != null)
            fr.remove();
        if (model != null)
            model.remove();
        if (partialDependence != null)
            partialDependence.remove();
    }
}

Example 30 with TwoDimTable

use of water.util.TwoDimTable in project h2o-3 by h2oai.

the class PartialDependenceTest method prostateBinaryPickCols.

@Test
public void prostateBinaryPickCols() {
    Frame fr = null;
    GBMModel model = null;
    PartialDependence partialDependence = null;
    try {
        // Frame
        fr = parse_test_file("smalldata/prostate/prostate.csv");
        for (String s : new String[] { "RACE", "GLEASON", "DPROS", "DCAPS", "CAPSULE" }) {
            Vec v = fr.remove(s);
            fr.add(s, v.toCategoricalVec());
            v.remove();
        }
        DKV.put(fr);
        // Model
        GBMModel.GBMParameters parms = new GBMModel.GBMParameters();
        parms._train = fr._key;
        parms._ignored_columns = new String[] { "ID" };
        parms._response_column = "CAPSULE";
        model = new GBM(parms).trainModel().get();
        // PartialDependence
        partialDependence = new PartialDependence(Key.<PartialDependence>make());
        //pick columns manually
        partialDependence._cols = new String[] { "DPROS", "GLEASON" };
        partialDependence._nbins = 10;
        partialDependence._model_id = (Key) model._key;
        partialDependence._frame_id = fr._key;
        partialDependence.execImpl().get();
        for (TwoDimTable t : partialDependence._partial_dependence_data) Log.info(t);
        Assert.assertTrue(partialDependence._partial_dependence_data.length == 2);
    } finally {
        if (fr != null)
            fr.remove();
        if (model != null)
            model.remove();
        if (partialDependence != null)
            partialDependence.remove();
    }
}