Packages

object EntropyAtomWeight extends Serializable

Linear Supertypes
Serializable, AnyRef, Any
Ordering
  1. Alphabetic
  2. By Inheritance
Inherited
  1. EntropyAtomWeight
  2. Serializable
  3. AnyRef
  4. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. Protected

Value Members

  1. def apply[A, B](genDist: FiniteDistribution[A], pDist: FiniteDistribution[B], qDist: FiniteDistribution[B], elem: B, initWeight: Double, hW: Double, klW: Double): EntropyAtomWeight

    More natural (but verbose) description of spire gradients for entropy tradeoff

    More natural (but verbose) description of spire gradients for entropy tradeoff

    A

    the type of objects of the initial distribution, e.g. terms

    B

    the type of objects of the final distribution, e.g. types

    genDist

    the initial generating distribution

    pDist

    the initial "theorem by proof" distribution

    qDist

    the initial "theorem by statement" distribution

    elem

    the lemma whose weight is learned

    initWeight

    the weight of the initial distribution in the final one

  2. def evolvedLemmaGens(ev: EvolvedStateLike, hW: Double, klW: Double): Vector[(Typ[Term], EntropyAtomWeight)]
  3. def evolvedLemmaIters(ev: EvolvedStateLike, hW: Double, klW: Double, sc: Double = 1, prune: Boolean = false): Vector[(Typ[Term], Iterator[Double])]
  4. def findGoal(init: TermState, base: TermState, tg: TermGenParams, steps: Int, maxDepth: Int, cutoff: Double, hW: Double, klW: Double, limit: FiniteDuration = 3.minutes, sc: Double = 1): Task[FiniteDistribution[Term]]
  5. def lemmaDist(ev: EvolvedStateLike, steps: Int, hW: Double, klW: Double, sc: Double = 1): FiniteDistribution[Term]
  6. def lemmaWeights(ev: EvolvedStateLike, steps: Int, hW: Double, klW: Double, sc: Double = 1): Vector[(Typ[Term], Double)]
  7. def proofWeightIter(ev: EvolvedStateLike, pf: Term, weight: Double, cutoff: Double, hW: Double, klW: Double, sc: Double = 1, prune: Boolean = false): Iterator[Double]
  8. def proofWeightTuned(ev: EvolvedStateLike, pf: Term, weight: Double, cutoff: Double, steps: Int, hW: Double, klW: Double, sc: Double = 1, prune: Boolean = false): Option[(Term, Double)]
  9. def tunedProofs(ev: EvolvedStateLike, pfs: Vector[(Term, Double)], cutoff: Double, steps: Int, hW: Double, klW: Double, sc: Double = 1, prune: Boolean = false): Vector[(Term, Double)]