data

Type Members

1. trait BatchFeaturizer[L, I, O] extends (Seq[Example[L, I]]) ⇒ Seq[Example[L, Seq[FeatureObservation[O]]]] with Serializable

   A function that converts a batch of objects of some input class into a sequence of FeatureObservations for an output class O.

2. class BinomialExampleIndexer extends (Example[(Int, Int), Seq[FeatureObservation[String]]]) ⇒ Example[(Int, Int), Seq[FeatureObservation[Int]]]

   Indexes the labels and features of a series of examples.

3. class BowFeaturizer extends Featurizer[String, String]

   A bag-of-words featurizer that simply tokenizes the input String by using whitespace and creates an observation for each token.

4. trait DataMatrix extends AnyRef

   A DataMatrix stores a double-valued label along with the double-valued features that go with it.

5. class ExactFeatureMap extends FeatureMap

   A feature map that stores all feature strings and their indices in an in-memory Map.

6. trait Example[+L, +T] extends Observation[T] with Labeled[L] with Serializable

   Represents a single example from a collection of data.

7. class ExampleIndexer extends (Example[String, Seq[FeatureObservation[String]]]) ⇒ Example[Int, Seq[FeatureObservation[Int]]]

   Indexes the labels and features of a series of examples.

8. trait FeatureMap extends Serializable

   A trait for classes that can index features represented as Strings.

9. case class FeatureObservation[F](feature: F, magnitude: Double = 1.0) extends Product with Serializable

   A feature with its observed magnitude in some context.

10. trait Featurizer[I, O] extends (I) ⇒ Seq[FeatureObservation[O]] with Serializable

    A function that converts objects of some input class into a sequence of FeatureObservations for an output class O.

11. class HashedExampleIndexer extends (Example[String, Seq[FeatureObservation[String]]]) ⇒ Example[Int, Seq[FeatureObservation[Int]]]

    Indexes the labels and features of a series of examples.

12. class HashedFeatureMap extends FeatureMap

    A feature map that uses the MurmurHash3 hash and mods on a prime giving the largest feature index that can be used.

13. trait LabelMap[L] extends Serializable

    A trait for classes that indexes labels and get labels of indexes.

14. trait Labeled[+L] extends AnyRef

    Something that has a label.

15. trait Multilabeled[L] extends Labeled[Set[L]]

    For any class that has one or more labels.

16. trait MultilabeledExample[L, +T] extends Example[Set[L], T] with Multilabeled[L]

    Represents a single example from a collection of data.

17. trait Observation[+T] extends Serializable

    Represents a single unlabeled example from a collection of data.

18. trait SparseFeatureDataset[Output] extends AnyRef

    Dataset of the form [featureIndex:featureValue*]

19. class TfidfBatchFeaturizer[L] extends BatchFeaturizer[L, String, String]

    A BatchFeaturizer that computes the tf-idf score of the terms in each Example.