com.cloudera.oryx.app.schema

Class InputSchema

java.lang.Object

com.cloudera.oryx.app.schema.InputSchema

All Implemented Interfaces:

Serializable

public final class InputSchema
extends Object
implements Serializable

Represents the essential schema information needed for some learning problems.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
InputSchema(com.typesafe.config.Config config)

Method Summary

Modifier and Type	Method and Description
int	featureToPredictorIndex(int featureIndex)
List<String>	getFeatureNames()
int	getNumFeatures()
int	getNumPredictors()
String	getTargetFeature()
int	getTargetFeatureIndex()
boolean	hasTarget()
boolean	isActive(int featureIndex)
boolean	isActive(String featureName)
boolean	isCategorical(int featureIndex)
boolean	isCategorical(String featureName)
boolean	isClassification()
boolean	isID(int featureIndex)
boolean	isID(String featureName)
boolean	isNumeric(int featureIndex)
boolean	isNumeric(String featureName)
boolean	isTarget(int featureIndex)
boolean	isTarget(String featureName)
int	predictorToFeatureIndex(int predictorIndex)
String	toString()

Methods inherited from class java.lang.Object

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

Constructor Detail

InputSchema

public InputSchema(com.typesafe.config.Config config)

Method Detail

getFeatureNames

public List<String> getFeatureNames()

Returns:

names of features in the schema, in order

getNumFeatures

public int getNumFeatures()

Returns:

total number of features described in the schema

getNumPredictors

public int getNumPredictors()

Returns:

number of features that are predictors -- not an ID, not ignored, not the target

isID

public boolean isID(String featureName)

Parameters:

featureName - feature name

Returns:

true iff feature is an ID feature

isID

public boolean isID(int featureIndex)

Parameters:

featureIndex - feature index

Returns:

true iff feature is an ID feature

isActive

public boolean isActive(String featureName)

Parameters:

featureName - feature name

Returns:

true iff feature is active -- not an ID or ignored

isActive

public boolean isActive(int featureIndex)

Parameters:

featureIndex - feature index

Returns:

true iff feature is active -- not an ID or ignored

isNumeric

public boolean isNumeric(String featureName)

Parameters:

featureName - feature name

Returns:

true iff feature is numeric

isNumeric

public boolean isNumeric(int featureIndex)

Parameters:

featureIndex - feature index

Returns:

true iff feature is a numeric

isCategorical

public boolean isCategorical(String featureName)

Parameters:

featureName - feature name

Returns:

true iff feature is categorical

isCategorical

public boolean isCategorical(int featureIndex)

Parameters:

featureIndex - feature index

Returns:

true iff feature is a categorical

isTarget

public boolean isTarget(String featureName)

Parameters:

featureName - feature name

Returns:

true iff the feature is the target

isTarget

public boolean isTarget(int featureIndex)

Parameters:

featureIndex - feature index

Returns:

true iff the feature is the target

getTargetFeature

public String getTargetFeature()

Returns:

name of feature that is the target

Throws:

IllegalStateException - if there is no target

getTargetFeatureIndex

public int getTargetFeatureIndex()

Returns:

index of feature that is the target

Throws:

IllegalStateException - if there is no target

hasTarget

public boolean hasTarget()

Returns:

true iff schema defines a target feature

isClassification

public boolean isClassification()

Returns:

true iff the target feature is categorical

Throws:

IllegalStateException - if there is no target

featureToPredictorIndex

public int featureToPredictorIndex(int featureIndex)

Parameters:

featureIndex - index (0-based) among all features

Returns:

index (0-based) among only predictors (not ID, not ignored, not target)

predictorToFeatureIndex

public int predictorToFeatureIndex(int predictorIndex)

Parameters:

predictorIndex - index (0-based) among only predictors (not ID, not ignored, not target)

Returns:

index (0-based) among all features

toString

public String toString()

Overrides:

toString in class Object