Serialized Form

Package hex.genmodel

Class hex.genmodel.GenModel extends java.lang.Object implements Serializable

Serialized Fields

_names

java.lang.String[] _names

Column names; last is response for supervised models

_domains

java.lang.String[][] _domains

Categorical (factor/enum) mappings, per column. Null for non-enum cols. Columns match the post-init cleanup columns. The last column holds the response col enums for SupervisedModels.

_offsetColumn

java.lang.String _offsetColumn

Name of the column with offsets (used for certain types of models).

Class hex.genmodel.GenMunger extends java.lang.Object implements Serializable

Serialized Fields

_steps

GenMunger.Step<T>[] _steps

Class hex.genmodel.GenMunger.Step extends java.lang.Object implements Serializable

Serialized Fields

_names

java.lang.String[] _names

_types

java.lang.String[] _types

_outNames

java.lang.String[] _outNames

_params

java.util.HashMap<K,V> _params

Class hex.genmodel.MojoModel extends GenModel implements Serializable

Serialized Fields

_category

ModelCategory _category

_uuid

java.lang.String _uuid

_supervised

boolean _supervised

_nfeatures

int _nfeatures

_nclasses

int _nclasses

_balanceClasses

boolean _balanceClasses

_defaultThreshold

double _defaultThreshold

_priorClassDistrib

double[] _priorClassDistrib

_modelClassDistrib

double[] _modelClassDistrib

Package hex.genmodel.algos

Class hex.genmodel.algos.DrfModel extends TreeBasedModel implements Serializable

Serialized Fields

_effective_n_classes

int _effective_n_classes

_binomial_double_trees

boolean _binomial_double_trees

Class hex.genmodel.algos.GbmModel extends TreeBasedModel implements Serializable

Serialized Fields

_family

DistributionFamily _family

_init_f

double _init_f

Class hex.genmodel.algos.TreeBasedModel extends MojoModel implements Serializable

Serialized Fields

_ntrees

int _ntrees

_compressed_trees

byte[][] _compressed_trees

Package hex.genmodel.easy

Class hex.genmodel.easy.EasyPredictModelWrapper extends java.lang.Object implements Serializable

Serialized Fields

m

GenModel m

modelColumnNameToIndexMap

java.util.HashMap<K,V> modelColumnNameToIndexMap

domainMap

java.util.HashMap<K,V> domainMap

convertUnknownCategoricalLevelsToNa

boolean convertUnknownCategoricalLevelsToNa

unknownCategoricalLevelsSeenPerColumn

java.util.concurrent.ConcurrentHashMap<K,V> unknownCategoricalLevelsSeenPerColumn

Class hex.genmodel.easy.RowData extends java.util.HashMap<java.lang.String,java.lang.Object> implements Serializable

Package hex.genmodel.easy.exception

Class hex.genmodel.easy.exception.PredictException extends java.lang.Exception implements Serializable

Class hex.genmodel.easy.exception.PredictUnknownCategoricalLevelException extends PredictException implements Serializable

Serialized Fields

columnName

java.lang.String columnName

unknownLevel

java.lang.String unknownLevel

Class hex.genmodel.easy.exception.PredictUnknownTypeException extends PredictException implements Serializable

Class hex.genmodel.easy.exception.PredictWrongModelCategoryException extends PredictException implements Serializable

Package hex.genmodel.easy.prediction

Class hex.genmodel.easy.prediction.AbstractPrediction extends java.lang.Object implements Serializable

Class hex.genmodel.easy.prediction.AutoEncoderModelPrediction extends AbstractPrediction implements Serializable

Class hex.genmodel.easy.prediction.BinomialModelPrediction extends AbstractPrediction implements Serializable

Serialized Fields

labelIndex

int labelIndex

0 or 1.

label

java.lang.String label

Label of the predicted level.

classProbabilities

double[] classProbabilities

This array of length two has the class probability for each class (aka categorical or factor level) in the response column. The array corresponds to the level names returned by:

 model.getDomainValues(model.getResponseIdx())
 

"Domain" is the internal H2O term for level names. The values in this array may be Double.NaN, which means NA (this will happen with GLM, for example, if one of the input values for a new data point is NA). If they are valid numeric values, then they will sum up to 1.0.

Class hex.genmodel.easy.prediction.ClusteringModelPrediction extends AbstractPrediction implements Serializable

Serialized Fields

cluster

int cluster

Chosen cluster for this data point.

Class hex.genmodel.easy.prediction.DimReductionModelPrediction extends AbstractPrediction implements Serializable

Serialized Fields

dimensions

double[] dimensions

Class hex.genmodel.easy.prediction.MultinomialModelPrediction extends AbstractPrediction implements Serializable

Serialized Fields

labelIndex

int labelIndex

Index number of the predicted class (aka categorical or factor level) in the response column.

label

java.lang.String label

Label of the predicted level.

classProbabilities

double[] classProbabilities

This array has an element for each class (aka categorical or factor level) in the response column. The array corresponds to the level names returned by:

 model.getDomainValues(model.getResponseIdx())
 

"Domain" is the internal H2O term for level names. The values in this array may be Double.NaN, which means NA. If they are valid numeric values, then they will sum up to 1.0.

Class hex.genmodel.easy.prediction.RegressionModelPrediction extends AbstractPrediction implements Serializable

Serialized Fields

value

double value

This value may be Double.NaN, which means NA (this will happen with GLM, for example, if one of the input values for a new data point is NA).