com.linkedin.feathr.common.value

Class FeatureValues

java.lang.Object

com.linkedin.feathr.common.value.FeatureValues

public class FeatureValues
extends java.lang.Object

Utility functions for constructing FeatureValue instances. User code that produces FeatureValues, such as Anchor extractors (in yet-to-come new Anchor extractor API), should prefer to use these when creating FeatureValues, when possible.

Method Summary

Modifier and Type	Method and Description
static BooleanFeatureValue	bool(boolean booleanValue) Returns a boolean feature value for a given boolean primitive value.
static BooleanFeatureValue	bool(TensorData tensor) Returns a boolean feature value for a given tensor, which must be a boolean-scalar.
static CategoricalFeatureValue	categorical(java.lang.String stringValue) Returns a categorical feature value for a given string categorical value.
static CategoricalFeatureValue	categorical(TensorData tensor) Returns a categorical feature for a given tensor value, assuming it is shaped as expected for a categorical feature.
static CategoricalSetFeatureValue	categoricalSet(java.util.Collection<java.lang.String> strings) Returns a categorical set feature for a given collection of strings.
static CategoricalSetFeatureValue	categoricalSet(java.util.Set<java.lang.String> strings) Returns a categorical set feature for a given set of strings.
static CategoricalSetFeatureValue	categoricalSet(java.lang.String... strings) Returns a categorical set feature for a given list of String arguments.
static CategoricalSetFeatureValue	categoricalSet(TensorData tensorData) Returns a categorical feature for a given tensor value, assuming it is shaped as expected for a categorical set feature.
static DenseVectorFeatureValue	denseVector(TensorData tensorData) Returns a dense-vector feature value for a given tensor, which must be a rank-1 tensor of value-type float.
static DenseVectorFeatureValue	denseVectorDynamicSize(float... floatArray) Returns a dense-vector feature value of unknown/dynamic size for a given list of float arguments.
static CategoricalSetFeatureValue	emptyCategoricalSet() Returns an empty categorical set.
static TermVectorFeatureValue	emptyTermVector() Returns an empty term-vector feature value.
static NumericFeatureValue	numeric(float floatValue) Returns a numeric feature value for a given float value
static NumericFeatureValue	numeric(TensorData tensor) Returns a numeric feature value from a given tensor, which must be a numeric-scalar
static TermVectorFeatureValue	termVector(java.util.Map<java.lang.String,java.lang.Float> termVector) Returns a term-vector feature value for a given term-vector map.
static TermVectorFeatureValue	termVector(java.lang.String t0, float v0) Returns a term-vector feature value with a single mapping, given by the term and value arguments.
static TermVectorFeatureValue	termVector(java.lang.String t0, float v0, java.lang.String t1, float v1) Returns a term-vector feature value with 2 mappings, given by the term and value arguments.
static TermVectorFeatureValue	termVector(java.lang.String t0, float v0, java.lang.String t1, float v1, java.lang.String t2, float v2) Returns a term-vector feature value with 3 mappings, given by the term and value arguments.
static TermVectorFeatureValue	termVector(TensorData tensorData) Returns a term-vector feature value for a given tensor, which must be rank-1 and have numeric value type.

Methods inherited from class java.lang.Object

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

Method Detail

numeric

public static NumericFeatureValue numeric(float floatValue)

Returns a numeric feature value for a given float value

numeric

public static NumericFeatureValue numeric(TensorData tensor)

Returns a numeric feature value from a given tensor, which must be a numeric-scalar

bool

public static BooleanFeatureValue bool(boolean booleanValue)

Returns a boolean feature value for a given boolean primitive value. (We would have named this function "boolean" but that is not allowed since it's a Java keyword.)

bool

public static BooleanFeatureValue bool(TensorData tensor)

Returns a boolean feature value for a given tensor, which must be a boolean-scalar. (We would have named this function "boolean" but that is not allowed since it's a Java keyword.)

categorical

public static CategoricalFeatureValue categorical(java.lang.String stringValue)

Returns a categorical feature value for a given string categorical value.

categorical

public static CategoricalFeatureValue categorical(TensorData tensor)

Returns a categorical feature for a given tensor value, assuming it is shaped as expected for a categorical feature.

categoricalSet

public static CategoricalSetFeatureValue categoricalSet(java.util.Collection<java.lang.String> strings)

Returns a categorical set feature for a given collection of strings.

categoricalSet

public static CategoricalSetFeatureValue categoricalSet(java.util.Set<java.lang.String> strings)

Returns a categorical set feature for a given set of strings.

categoricalSet

public static CategoricalSetFeatureValue categoricalSet(java.lang.String... strings)

Returns a categorical set feature for a given list of String arguments.

emptyCategoricalSet

public static CategoricalSetFeatureValue emptyCategoricalSet()

Returns an empty categorical set.

categoricalSet

public static CategoricalSetFeatureValue categoricalSet(TensorData tensorData)

Returns a categorical feature for a given tensor value, assuming it is shaped as expected for a categorical set feature.

denseVectorDynamicSize

public static DenseVectorFeatureValue denseVectorDynamicSize(float... floatArray)

Returns a dense-vector feature value of unknown/dynamic size for a given list of float arguments.

denseVector

public static DenseVectorFeatureValue denseVector(TensorData tensorData)

Returns a dense-vector feature value for a given tensor, which must be a rank-1 tensor of value-type float.

termVector

public static TermVectorFeatureValue termVector(java.util.Map<java.lang.String,java.lang.Float> termVector)

Returns a term-vector feature value for a given term-vector map.

termVector

public static TermVectorFeatureValue termVector(TensorData tensorData)

Returns a term-vector feature value for a given tensor, which must be rank-1 and have numeric value type.

termVector

public static TermVectorFeatureValue termVector(java.lang.String t0,
                                                float v0)

Returns a term-vector feature value with a single mapping, given by the term and value arguments.

termVector

public static TermVectorFeatureValue termVector(java.lang.String t0,
                                                float v0,
                                                java.lang.String t1,
                                                float v1)

Returns a term-vector feature value with 2 mappings, given by the term and value arguments.

termVector

public static TermVectorFeatureValue termVector(java.lang.String t0,
                                                float v0,
                                                java.lang.String t1,
                                                float v1,
                                                java.lang.String t2,
                                                float v2)

Returns a term-vector feature value with 3 mappings, given by the term and value arguments.

emptyTermVector

public static TermVectorFeatureValue emptyTermVector()

Returns an empty term-vector feature value.