hex

Class ModelMetricsRegression

java.lang.Object

water.Iced<T>

water.Keyed<ModelMetrics>

hex.ModelMetrics

hex.ModelMetricsSupervised

hex.ModelMetricsRegression

All Implemented Interfaces:

java.io.Externalizable, java.io.Serializable, java.lang.Cloneable, Freezable<ModelMetrics>

Direct Known Subclasses:

ModelMetricsRegressionGLM

public class ModelMetricsRegression
extends ModelMetricsSupervised

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	ModelMetricsRegression.MetricBuilderRegression<T extends ModelMetricsRegression.MetricBuilderRegression<T>>

Nested classes/interfaces inherited from class hex.ModelMetricsSupervised

ModelMetricsSupervised.MetricBuilderSupervised<T extends ModelMetricsSupervised.MetricBuilderSupervised<T>>

Nested classes/interfaces inherited from class hex.ModelMetrics

ModelMetrics.MetricBuilder<T extends ModelMetrics.MetricBuilder<T>>

Field Summary

Fields

Modifier and Type	Field and Description
double	_mean_absolute_error
double	_mean_residual_deviance
double	_root_mean_squared_log_error

Fields inherited from class hex.ModelMetricsSupervised

_domain, _sigma

Fields inherited from class hex.ModelMetrics

_description, _MSE, _nobs, _scoring_time

Fields inherited from class water.Keyed

_key

Constructor Summary

Constructors

Constructor and Description
ModelMetricsRegression(Model model, Frame frame, long nobs, double mse, double sigma, double mae, double rmsle, double meanResidualDeviance)

Method Summary

Methods

Modifier and Type	Method and Description
static double	computeHuberDelta(Vec actual, Vec preds, Vec weight, double huberAlpha)
static ModelMetricsRegression	getFromDKV(Model model, Frame frame)
double	mae()
static ModelMetricsRegression	make(Vec predicted, Vec actual, hex.genmodel.utils.DistributionFamily family) Build a Regression ModelMetrics object from predicted and actual targets
double	mean_residual_deviance()
double	residual_deviance() For all algos except GLM this is mean residual deviance.
double	rmsle()
java.lang.String	toString()

Methods inherited from class hex.ModelMetricsSupervised

r2

Methods inherited from class hex.ModelMetrics

auc_obj, buildKey, calcVarImp, calcVarImp, calcVarImp, calcVarImp, checksum_impl, cm, deepCloneWithDifferentModelAndFrame, frame, getAllowedMetrics, getMetricFromModel, getMetricFromModelMetric, hr, isForFrame, isForModel, model, mse, residual_degrees_of_freedom, rmse, sortModelsByMetric, sortModelsByMetric

Methods inherited from class water.Keyed

checksum, makeSchema, readAll_impl, readAll, remove_impl, remove, remove, remove, remove, writeAll_impl, writeAll

Methods inherited from class water.Iced

asBytes, clone, copyOver, frozenType, read, readExternal, readJSON, reloadFromBytes, toJsonString, write, writeExternal, writeJSON

Methods inherited from class java.lang.Object

equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

_mean_residual_deviance

public final double _mean_residual_deviance

_mean_absolute_error

public final double _mean_absolute_error

_root_mean_squared_log_error

public final double _root_mean_squared_log_error

Constructor Detail

ModelMetricsRegression

public ModelMetricsRegression(Model model,
                      Frame frame,
                      long nobs,
                      double mse,
                      double sigma,
                      double mae,
                      double rmsle,
                      double meanResidualDeviance)

Method Detail

residual_deviance

public double residual_deviance()

For all algos except GLM this is mean residual deviance. For GLM it's total residual deviance.

mean_residual_deviance

public double mean_residual_deviance()

mae

public double mae()

rmsle

public double rmsle()

getFromDKV

public static ModelMetricsRegression getFromDKV(Model model,
                                Frame frame)

toString

public java.lang.String toString()

Overrides:

toString in class ModelMetricsSupervised

make

public static ModelMetricsRegression make(Vec predicted,
                          Vec actual,
                          hex.genmodel.utils.DistributionFamily family)

Build a Regression ModelMetrics object from predicted and actual targets

Parameters:

predicted - A Vec containing predicted values

actual - A Vec containing the actual target values

Returns:

ModelMetrics object

computeHuberDelta

public static double computeHuberDelta(Vec actual,
                       Vec preds,
                       Vec weight,
                       double huberAlpha)