com.cra.figaro.algorithm.factored

factors

package factors

Visibility
  1. Public
  2. All

Type Members

  1. class BasicFactor[T] extends Factor[T]

    Default implementation of Factor.

    Default implementation of Factor.

    A factor is associated with a set of variables and specifies a value for every combination of assignments to those variables. Factors are parameterized by the types of values they contain.

  2. case class BooleanSemiring() extends Semiring[Boolean] with Product with Serializable

  3. case class BoundsSumProductSemiring() extends DivideableSemiRing[(Double, Double)] with Product with Serializable

    Semiring for computing sums and products with lower and upper bounds.

  4. class ConditionalSelector[T] extends BasicFactor[T]

    ConditionalSelector Factor.

    ConditionalSelector Factor. A conditional selector factor is associated chain elements and represents the relationship between parent elements and outcome elements. It is a special form of sparse factor where the default element is not 0 but 1 (don't care) which is the most frequent in this type of factor.

  5. class ConstraintFactor[T] extends BasicFactor[T]

    Constraint Factor.

    Constraint Factor. A factor is associated with a set of variables and specifies a value for every combination of assignments to those variables. Factors are parameterized by the types of values they contain.

    This factor is used to model constraints on elements.

  6. trait DivideableSemiRing[T] extends Semiring[T]

  7. class ElementVariable[T] extends Variable[T]

    Variables generated from elements.

  8. trait Factor[T] extends AnyRef

    Definition of Factor

    Definition of Factor

    A factor is associated with a set of variables and specifies a value for every combination of assignments to those variables.

    Factors are parameterized by the type of the Variables they contain and contain a semiring that defines the mathematical operation to be performed on the values Parent variables are distinguished from the output variable.

    Refactored by

  9. trait FactorMaker[T] extends AnyRef

    A trait for elements that are able to construct their own Factor.

  10. class Indices extends Iterable[List[Int]]

  11. class InternalVariable[T] extends Variable[T]

    Variables that are internal to Factors.

    Variables that are internal to Factors.

    This is the same as a temporary variable, but is more explicitly identified

  12. case class LogMaxProductSemiring() extends DivideableSemiRing[Double] with Product with Serializable

    Semiring for computing maxs and products with log probabilities.

  13. case class LogSumProductSemiring() extends DivideableSemiRing[Double] with Product with Serializable

    Semiring for computing sums and products with log probabilities.

  14. case class MaxProductSemiring() extends DivideableSemiRing[Double] with Product with Serializable

  15. class ParameterizedVariable[T] extends ElementVariable[T]

    Variables generated from parameterized elements

  16. trait Semiring[T] extends AnyRef

    Operations in factored algorithms are defined by a semiring algebraic structure.

    Operations in factored algorithms are defined by a semiring algebraic structure. Each semiring defines a product and sum operation, and a value for zero and one which satisfy a set of properties. Different semirings are appropriate for certain algorithms and data types.

  17. class SparseFactor[T] extends BasicFactor[T]

    Sparse implementation of Factor.

    Sparse implementation of Factor.

    Factors are parameterized by the type of the Variables they contain and contain a semiring that defines the mathematical operation to be performed on the values Parent variables are distinguished from the output variable.

    This implementation stores only non-default (non-zero) elements and supplies special sum and product methods to account for the missing values.

  18. class SufficientStatisticsSemiring extends Semiring[(Double, Map[Parameter[_], Seq[Double]])]

    Sum and product operations defined for sufficient statistics.

    Sum and product operations defined for sufficient statistics. Statistics consist of a probability and counts of the number of times various values have been seen.

  19. case class SumProductSemiring() extends DivideableSemiRing[Double] with Product with Serializable

  20. case class SumProductUtilitySemiring() extends DivideableSemiRing[(Double, Double)] with Product with Serializable

  21. class Variable[T] extends AnyRef

    Variables that appear in factors.

Value Members

  1. object Factor

  2. object Factory

    Methods for creating probabilistic factors associated with elements.

  3. object SufficientStatisticsSemiring

  4. object Variable

  5. package factory

Ungrouped