hex.genmodel.algos.glm.GlmMojoModelBase Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of h2o-genmodel Show documentation

H2O GenModel

There is a newer version: 3.46.0.6

package hex.genmodel.algos.glm;

import hex.genmodel.MojoModel;

public abstract class GlmMojoModelBase extends MojoModel {

  boolean _useAllFactorLevels;

  int _cats;
  int[] _catModes;
  int[] _catOffsets;

  int _nums;
  double[] _numMeans;
  boolean _meanImputation;

  double[] _beta;

  String _family;
  boolean _versionSupportOffset;
  
  double _dispersion_estimated;

  GlmMojoModelBase(String[] columns, String[][] domains, String responseColumn) {
    super(columns, domains, responseColumn);
  }

  public double[] getBeta() {
    return _beta;
  }
  
  public double getDispersionEstimated() {
    return _dispersion_estimated;
  }

  void init() {
    _versionSupportOffset = _mojo_version >= 1.1;
  }

  @Override
  public final double[] score0(double[] data, double[] preds) {
      return score0(data, 0, preds);
  }

  void imputeMissingWithMeans(double[] data) {
    for (int i = 0; i < _cats; ++i)
      if (Double.isNaN(data[i])) data[i] = _catModes[i];
    for (int i = 0; i < _nums; ++i)
      if (Double.isNaN(data[i + _cats])) data[i + _cats] = _numMeans[i];
  }

  @Override
  public String[] getOutputNames() {
    // special handling of binomial case where response domain is not represented
    if (nclasses() == 2 && getDomainValues(getResponseIdx()) == null) {
      return new String[]{"predict", "0", "1"};
    }
    return super.getOutputNames();
  }
}