Model.Output (h2o-core 3.42.0 API)

java.lang.Object
- water.Iced
- - hex.Model.Output

All Implemented Interfaces:

java.io.Externalizable, java.io.Serializable, java.lang.Cloneable, Freezable

Direct Known Subclasses:

ClusteringModel.ClusteringOutput, QuantileModel.QuantileOutput

Enclosing class:

Model<M extends Model<M,P,O>,P extends Model.Parameters,O extends Model.Output>
```
public abstract static class Model.Output
extends Iced
```
Model-specific output class. Each model sub-class contains an instance of one of these containing its "output": the pieces of the model needed for scoring. E.g. KMeansModel has a KMeansOutput extending Model.Output which contains the cluster centers. The output also includes the names, domains and other fields which are determined at training time.

See Also:

Serialized Form

Field Summary

Fields
Modifier and Type	Field and Description
`java.lang.String[]`	`_column_types`
`Key<Frame>`	`_cross_validation_fold_assignment_frame_id`
`Key<Frame>`	`_cross_validation_holdout_predictions_frame_id`
`ModelMetrics`	`_cross_validation_metrics` Cross-Validation metrics obtained during model training
`TwoDimTable`	`_cross_validation_metrics_summary` Summary of cross-validation metrics of all k-fold models
`Key[]`	`_cross_validation_models` List of Keys to cross-validation models (non-null iff _parms._nfolds > 1 or _parms._fold_column != null)
`Key[]`	`_cross_validation_predictions` List of Keys to cross-validation predictions (if requested)
`TwoDimTable[]`	`_cv_scoring_history`
`double`	`_defaultThreshold` Default threshold used to make decision about binomial predictions -1 if is not set by user - than the default threshold is 0.5 if metrics are not set (0, 1> custom default threshold or validation metric threshold or training metric threshold
`double[]`	`_distribution`
`java.lang.String[][]`	`_domains` Categorical/factor mappings, per column.
`long`	`_end_time`
`protected boolean`	`_hasFold`
`protected boolean`	`_hasOffset` The name of the response column (which is always the last column).
`protected boolean`	`_hasTreatment`
`protected boolean`	`_hasWeights`
`protected boolean`	`_isSupervised`
`Job`	`_job` Job info: final status (canceled, crashed), build time
`TwoDimTable`	`_model_summary` User-facing model summary - Display model type, complexity, size and other useful stats
`double[]`	`_modelClassDist`
`java.lang.String[]`	`_names` Columns used in the model and are used to match up with scoring data columns.
`double[]`	`_orig_projection_array`
`java.lang.String[][]`	`_origDomains`
`java.lang.String[]`	`_origNames`
`double[]`	`_priorClassDist`
`TwoDimTable[]`	`_reproducibility_information_table` Reproducibility information describing the current cluster configuration, each node configuration and checksums for each frame used on the input of the algorithm
`long`	`_run_time`
`TwoDimTable`	`_scoring_history` User-facing model scoring history - 2D table with modeling accuracy as a function of time/trees/epochs/iterations, etc.
`long`	`_start_time`
`long`	`_total_run_time`
`ModelMetrics`	`_training_metrics` Training set metrics obtained during model training
`ModelMetrics`	`_validation_metrics` Validation set metrics obtained during model training (if a validation data set was specified)

Constructor Summary

Constructors
Modifier	Constructor and Description
	`Output()`
	`Output(boolean hasWeights, boolean hasOffset, boolean hasFold)`
	`Output(ModelBuilder b)` Any final prep-work just before model-building starts, but after the user has clicked "go".
`protected`	`Output(ModelBuilder b, Frame train)`

Method Summary

All Methods Instance Methods Concrete Methods Deprecated Methods
Modifier and Type	Method and Description
`void`	`changeModelMetricsKey(Key modelkey)`
`protected long`	`checksum_impl()`
`java.lang.String[]`	`classNames()` Names of levels for a categorical response column.
`Key<ModelMetrics>[]`	`clearModelMetrics(boolean keepModelTrainingMetrics)`
`TwoDimTable`	`createInputFramesInformationTable(ModelBuilder modelBuilder)`
`double`	`defaultThreshold()`
`java.lang.String[]`	`features()` Returns features used by the model
`int`	`foldIdx()`
`java.lang.String`	`foldName()`
`int`	`getInformationTableNumRows()`
`hex.ModelCategory`	`getModelCategory()`
`Key<ModelMetrics>[]`	`getModelMetrics()`
`TwoDimTable`	`getVariableImportances()` Retrieves variable importances
`boolean`	`hasFold()`
`boolean`	`hasOffset()`
`boolean`	`hasResponse()`
`boolean`	`hasTreatment()`
`boolean`	`hasWeights()`
`Model.InteractionBuilder`	`interactionBuilder()`
`boolean`	`isAutoencoder()`
`boolean`	`isBinomialClassifier()` Is this model a binomial classification model? (v.
`boolean`	`isClassifier()` Is this model a classification model? (v.
`boolean`	`isMultinomialClassifier()` Is this model a multinomial classification model (supervised and nclasses() > 2
`boolean`	`isSupervised()`
`protected int`	`lastSpecialColumnIdx()`
`int`	`nclasses()` Number of classes in the response column if it is categorical and the model is supervised.
`int`	`nfeatures()` Returns number of input features (OK for most supervised methods, need to override for unsupervised!)
`int`	`offsetIdx()`
`java.lang.String`	`offsetName()`
`void`	`printTwoDimTables(java.lang.StringBuilder sb, java.lang.Object o)`
`void`	`resetThreshold(double value)`
`int`	`responseIdx()`
`java.lang.String`	`responseName()`
`void`	`setNames(java.lang.String[] names)` Deprecated. as of March 6, 2019, replaced by (@link #setNames(String[] names, String[] columnTypes))
`void`	`setNames(java.lang.String[] names, java.lang.String[] columntypes)`
`protected void`	`startClock()`
`protected void`	`stopClock()`
`java.lang.String`	`toString()`
`int`	`treatmentIdx()`
`int`	`weightsIdx()`
`java.lang.String`	`weightsName()`

Methods inherited from class water.Iced
asBytes, clone, copyOver, frozenType, read, readExternal, readJSON, reloadFromBytes, toJsonBytes, toJsonString, write, writeExternal, writeJSON

Methods inherited from class java.lang.Object
equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

- Field Detail
  - _names
```
public java.lang.String[] _names
```
    Columns used in the model and are used to match up with scoring data columns. The last name is the response column name (if any).
  - _column_types
```
public java.lang.String[] _column_types
```
  - _origNames
```
public java.lang.String[] _origNames
```
  - _domains
```
public java.lang.String[][] _domains
```
    Categorical/factor mappings, per column. Null for non-categorical cols. Columns match the post-init cleanup columns. The last column holds the response col categoricals for SupervisedModels.
  - _origDomains
```
public java.lang.String[][] _origDomains
```
  - _orig_projection_array
```
public double[] _orig_projection_array
```
  - _cross_validation_models
```
public Key[] _cross_validation_models
```
    List of Keys to cross-validation models (non-null iff _parms._nfolds > 1 or _parms._fold_column != null)
  - _cross_validation_predictions
```
public Key[] _cross_validation_predictions
```
    List of Keys to cross-validation predictions (if requested)
  - _cross_validation_holdout_predictions_frame_id
```
public Key<Frame> _cross_validation_holdout_predictions_frame_id
```
  - _cross_validation_fold_assignment_frame_id
```
public Key<Frame> _cross_validation_fold_assignment_frame_id
```
  - _start_time
```
public long _start_time
```
  - _end_time
```
public long _end_time
```
  - _run_time
```
public long _run_time
```
  - _total_run_time
```
public long _total_run_time
```
  - _job
```
public Job _job
```
    Job info: final status (canceled, crashed), build time
  - _training_metrics
```
public ModelMetrics _training_metrics
```
    Training set metrics obtained during model training
  - _validation_metrics
```
public ModelMetrics _validation_metrics
```
    Validation set metrics obtained during model training (if a validation data set was specified)
  - _cross_validation_metrics
```
public ModelMetrics _cross_validation_metrics
```
    Cross-Validation metrics obtained during model training
  - _cross_validation_metrics_summary
```
public TwoDimTable _cross_validation_metrics_summary
```
    Summary of cross-validation metrics of all k-fold models
  - _model_summary
```
public TwoDimTable _model_summary
```
    User-facing model summary - Display model type, complexity, size and other useful stats
  - _reproducibility_information_table
```
public TwoDimTable[] _reproducibility_information_table
```
    Reproducibility information describing the current cluster configuration, each node configuration and checksums for each frame used on the input of the algorithm
  - _scoring_history
```
public TwoDimTable _scoring_history
```
    User-facing model scoring history - 2D table with modeling accuracy as a function of time/trees/epochs/iterations, etc.
  - _cv_scoring_history
```
public TwoDimTable[] _cv_scoring_history
```
  - _distribution
```
public double[] _distribution
```
  - _modelClassDist
```
public double[] _modelClassDist
```
  - _priorClassDist
```
public double[] _priorClassDist
```
  - _isSupervised
```
protected boolean _isSupervised
```
  - _defaultThreshold
```
public double _defaultThreshold
```
    Default threshold used to make decision about binomial predictions -1 if is not set by user - than the default threshold is 0.5 if metrics are not set (0, 1> custom default threshold or validation metric threshold or training metric threshold
  - _hasOffset
```
protected boolean _hasOffset
```
    The name of the response column (which is always the last column).
  - _hasWeights
```
protected boolean _hasWeights
```
  - _hasFold
```
protected boolean _hasFold
```
  - _hasTreatment
```
protected boolean _hasTreatment
```
- Constructor Detail
  - Output
```
public Output()
```
  - Output
```
public Output(boolean hasWeights,
              boolean hasOffset,
              boolean hasFold)
```
  - Output
```
public Output(ModelBuilder b)
```
    Any final prep-work just before model-building starts, but after the user has clicked "go". E.g., converting a response column to an categorical touches the entire column (can be expensive), makes a parallel vec (Key/Data leak management issues), and might throw IAE if there are too many classes.
  - Output
```
protected Output(ModelBuilder b,
                 Frame train)
```
- Method Detail
  - setNames
```
@Deprecated
public void setNames(java.lang.String[] names)
```
    Deprecated. as of March 6, 2019, replaced by (@link #setNames(String[] names, String[] columnTypes))
  - setNames
```
public void setNames(java.lang.String[] names,
                     java.lang.String[] columntypes)
```
  - startClock
```
protected void startClock()
```
  - stopClock
```
protected void stopClock()
```
  - nfeatures
```
public int nfeatures()
```
    Returns number of input features (OK for most supervised methods, need to override for unsupervised!)
  - features
```
public java.lang.String[] features()
```
    Returns features used by the model
  - isSupervised
```
public boolean isSupervised()
```
  - hasOffset
```
public boolean hasOffset()
```
  - hasWeights
```
public boolean hasWeights()
```
  - hasFold
```
public boolean hasFold()
```
  - hasTreatment
```
public boolean hasTreatment()
```
  - hasResponse
```
public boolean hasResponse()
```
  - responseName
```
public java.lang.String responseName()
```
  - weightsName
```
public java.lang.String weightsName()
```
  - offsetName
```
public java.lang.String offsetName()
```
  - foldName
```
public java.lang.String foldName()
```
  - interactionBuilder
```
public Model.InteractionBuilder interactionBuilder()
```
  - lastSpecialColumnIdx
```
protected int lastSpecialColumnIdx()
```
  - weightsIdx
```
public int weightsIdx()
```
  - offsetIdx
```
public int offsetIdx()
```
  - foldIdx
```
public int foldIdx()
```
  - responseIdx
```
public int responseIdx()
```
  - treatmentIdx
```
public int treatmentIdx()
```
  - classNames
```
public java.lang.String[] classNames()
```
    Names of levels for a categorical response column.
  - isClassifier
```
public boolean isClassifier()
```
    Is this model a classification model? (v. a regression or clustering model)
  - isBinomialClassifier
```
public boolean isBinomialClassifier()
```
    Is this model a binomial classification model? (v. a regression or clustering model)
  - isMultinomialClassifier
```
public boolean isMultinomialClassifier()
```
    Is this model a multinomial classification model (supervised and nclasses() > 2
  - nclasses
```
public int nclasses()
```
    Number of classes in the response column if it is categorical and the model is supervised.
  - getModelCategory
```
public hex.ModelCategory getModelCategory()
```
  - isAutoencoder
```
public boolean isAutoencoder()
```
  - getVariableImportances
```
public TwoDimTable getVariableImportances()
```
    Retrieves variable importances
    
    Returns:
    
    instance of TwoDimTable if model supports variable importances, null otherwise
  - clearModelMetrics
```
public Key<ModelMetrics>[] clearModelMetrics(boolean keepModelTrainingMetrics)
```
  - getModelMetrics
```
public Key<ModelMetrics>[] getModelMetrics()
```
  - changeModelMetricsKey
```
public void changeModelMetricsKey(Key modelkey)
```
  - checksum_impl
```
protected long checksum_impl()
```
  - defaultThreshold
```
public double defaultThreshold()
```
  - resetThreshold
```
public void resetThreshold(double value)
```
  - printTwoDimTables
```
public void printTwoDimTables(java.lang.StringBuilder sb,
                              java.lang.Object o)
```
  - toString
```
public java.lang.String toString()
```
    Overrides:
    
    toString in class java.lang.Object
  - createInputFramesInformationTable
```
public TwoDimTable createInputFramesInformationTable(ModelBuilder modelBuilder)
```
  - getInformationTableNumRows
```
public int getInformationTableNumRows()
```

Class Model.Output

Field Summary

Constructor Summary

Method Summary

Methods inherited from class water.Iced

Methods inherited from class java.lang.Object

Field Detail

_names

_column_types

_origNames

_domains

_origDomains

_orig_projection_array

_cross_validation_models

_cross_validation_predictions

_cross_validation_holdout_predictions_frame_id

_cross_validation_fold_assignment_frame_id

_start_time

_end_time

_run_time

_total_run_time

_job

_training_metrics

_validation_metrics

_cross_validation_metrics

_cross_validation_metrics_summary

_model_summary

_reproducibility_information_table

_scoring_history

_cv_scoring_history

_distribution

_modelClassDist

_priorClassDist

_isSupervised

_defaultThreshold

_hasOffset

_hasWeights

_hasFold

_hasTreatment

Constructor Detail

Output

Output

Output

Output

Method Detail

setNames

setNames

startClock

stopClock

nfeatures

features

isSupervised

hasOffset

hasWeights

hasFold

hasTreatment

hasResponse

responseName

weightsName

offsetName

foldName

interactionBuilder

lastSpecialColumnIdx

weightsIdx

offsetIdx

foldIdx

responseIdx

treatmentIdx

classNames

isClassifier

isBinomialClassifier

isMultinomialClassifier

nclasses

getModelCategory

isAutoencoder

getVariableImportances

clearModelMetrics

getModelMetrics

changeModelMetricsKey

checksum_impl