Configuration Options
Key: admmmemorytermstore.internalstore
Type: String
Default Value: org.linqs.psl.reasoner.term.MemoryTermStore
Module: psl-core
Defining Class: org.linqs.psl.reasoner.admm.term.ADMMTermStore
Description: Initial size for the memory store.
Key: admmreasoner.epsilonabs
Type: float
Default Value: 1e-5f
Module: psl-core
Defining Class: org.linqs.psl.reasoner.admm.ADMMReasoner
Description: Absolute error component of stopping criteria. Should be positive.
Key: admmreasoner.epsilonrel
Type: float
Default Value: 1e-3f
Module: psl-core
Defining Class: org.linqs.psl.reasoner.admm.ADMMReasoner
Description: Relative error component of stopping criteria. Should be positive.
Key: admmreasoner.initialconsensusvalue
Type: String
Default Value: InitialValue.RANDOM.toString()
Module: psl-core
Defining Class: org.linqs.psl.reasoner.admm.ADMMReasoner
Description: The starting value for consensus variables. Values should come from the InitialValue enum.
Key: admmreasoner.initiallocalvalue
Type: String
Default Value: InitialValue.RANDOM.toString()
Module: psl-core
Defining Class: org.linqs.psl.reasoner.admm.ADMMReasoner
Description: The starting value for local variables. Values should come from the InitialValue enum.
Key: admmreasoner.maxiterations
Type: int
Default Value: 25000
Module: psl-core
Defining Class: org.linqs.psl.reasoner.admm.ADMMReasoner
Description: The maximum number of iterations of ADMM to perform in a round of inference.
Key: admmreasoner.objectivebreak
Type: boolean
Default Value: true
Module: psl-core
Defining Class: org.linqs.psl.reasoner.admm.ADMMReasoner
Description: Stop if the objective has not changed since the last logging period (see LOG_PERIOD).
Key: admmreasoner.stepsize
Type: float
Default Value: 1.0f
Module: psl-core
Defining Class: org.linqs.psl.reasoner.admm.ADMMReasoner
Description: Step size. Higher values result in larger steps. Should be positive.
Key: admmtermgenerator.invertnegativeweights
Type: boolean
Default Value: false
Module: psl-core
Defining Class: org.linqs.psl.reasoner.admm.term.ADMMTermGenerator
Description: If true, then invert negative weight rules into their positive weight counterparts (negate the weight and expression).
Key: arithmeticrule.delim
Type: String
Default Value: ;
Module: psl-core
Defining Class: org.linqs.psl.model.rule.arithmetic.AbstractArithmeticRule
Description: The delimiter to use when building summation substitutions. Make sure the value for this key does not appear in ground atoms that use a summation.
Key: booleanmaxwalksat.maxflips
Type: int
Default Value: 50000
Module: psl-core
Defining Class: org.linqs.psl.reasoner.bool.BooleanMaxWalkSat
Description: Key for positive integer property that is the maximum number of flips to try during optimization
Key: booleanmaxwalksat.noise
Type: double
Default Value: 0.01
Module: psl-core
Defining Class: org.linqs.psl.reasoner.bool.BooleanMaxWalkSat
Description: Key for double property in [0,1] that is the probability of randomly perturbing an atom in a randomly chosen potential
Key: booleanmcsat.numburnin
Type: int
Default Value: 500
Module: psl-core
Defining Class: org.linqs.psl.reasoner.bool.BooleanMCSat
Description: Number of burn-in samples
Key: booleanmcsat.numsamples
Type: int
Default Value: 2500
Module: psl-core
Defining Class: org.linqs.psl.reasoner.bool.BooleanMCSat
Description: Key for length of Markov chain
Key: categoricalevaluator.categoryindexes
Type:
Default Value:
Module: psl-core
Defining Class: org.linqs.psl.evaluation.statistics.CategoricalEvaluator
Description: The index of the arguments in the predicate (delimited by colons).
Key: categoricalevaluator.defaultpredicate
Type:
Default Value:
Module: psl-core
Defining Class: org.linqs.psl.evaluation.statistics.CategoricalEvaluator
Description: The default predicate to use when none are supplied.
Key: categoricalevaluator.representative
Type:
Default Value:
Module: psl-core
Defining Class: org.linqs.psl.evaluation.statistics.CategoricalEvaluator
Description: The representative metric. Default to accuracy. Must match a string from the RepresentativeMetric enum.
Key: continuousevaluator.representative
Type:
Default Value:
Module: psl-core
Defining Class: org.linqs.psl.evaluation.statistics.ContinuousEvaluator
Description: The representative metric. Default to MSE. Must match a string from the RepresentativeMetric enum.
Key: continuousrandomgridsearch.maxlocations
Type: int
Default Value: 250
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.grid.ContinuousRandomGridSearch
Description: The max number of locations to search.
Key: discreteevaluator.representative
Type:
Default Value:
Module: psl-core
Defining Class: org.linqs.psl.evaluation.statistics.DiscreteEvaluator
Description: The representative metric. Default to F1. Must match a string from the RepresentativeMetric enum.
Key: discreteevaluator.threshold
Type:
Default Value:
Module: psl-core
Defining Class: org.linqs.psl.evaluation.statistics.DiscreteEvaluator
Description: The truth threshold.
Key: em.iterations
Type: int
Default Value: 10
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.em.ExpectationMaximization
Description: Key for positive int property for the number of iterations of expectation maximization to perform
Key: em.tolerance
Type: double
Default Value: 1e-3
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.em.ExpectationMaximization
Description: Key for positive double property for the minimum absolute change in weights such that EM is considered converged
Key: executablereasoner.cleanupinput
Type: boolean
Default Value: true
Module: psl-core
Defining Class: org.linqs.psl.reasoner.ExecutableReasoner
Description: Key for boolean property for whether to delete the input file to external the reasoner on close.
Key: executablereasoner.cleanupoutput
Type: boolean
Default Value: true
Module: psl-core
Defining Class: org.linqs.psl.reasoner.ExecutableReasoner
Description: Key for boolean property for whether to delete the output file to external the reasoner on close.
Key: executablereasoner.executablepath
Type:
Default Value:
Module: psl-core
Defining Class: org.linqs.psl.reasoner.ExecutableReasoner
Description: Key for int property for the path of the executable.
Key: gridsearch.weights
Type: String
Default Value: 0.001:0.01:0.1:1:10
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.grid.GridSearch
Description: A comma-separated list of possible weights. These weights should be in some sorted order.
Key: guidedrandomgridsearch.explorelocations
Type: int
Default Value: 10
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.grid.GuidedRandomGridSearch
Description: The number of initial seed locations to explore based off of whichever ones score the best.
Key: guidedrandomgridsearch.seedlocations
Type: int
Default Value: 25
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.grid.GuidedRandomGridSearch
Description: The number of locations to initially search.
Key: hardem.adagrad
Type: boolean
Default Value: false
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.em.HardEM
Description: Key for Boolean property that indicates whether to use AdaGrad subgradient scaling, the adaptive subgradient algorithm of John Duchi, Elad Hazan, Yoram Singer (JMLR 2010). If TRUE, will override other step scheduling options (but not scaling).
Key: hyperband.basebracketsize
Type: int
Default Value: 10
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.Hyperband
Description: The base number of weight configurations for each brackets.
Key: hyperband.numbrackets
Type: int
Default Value: 4
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.Hyperband
Description: The number of brackets to consider. This is computed in vanilla Hyperband.
Key: hyperband.survival
Type: int
Default Value: 4
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.Hyperband
Description: The proportion of configs that survive each round in a brancket.
Key: inference.groundrulestore
Type: String
Default Value: org.linqs.psl.application.groundrulestore.MemoryGroundRuleStore
Module: psl-core
Defining Class: org.linqs.psl.application.inference.InferenceApplication
Description: The class to use for ground rule storage.
Key: inference.reasoner
Type: String
Default Value: org.linqs.psl.reasoner.admm.ADMMReasoner
Module: psl-core
Defining Class: org.linqs.psl.application.inference.InferenceApplication
Description: The class to use for a reasoner.
Key: inference.termgenerator
Type: String
Default Value: org.linqs.psl.reasoner.admm.term.ADMMTermGenerator
Module: psl-core
Defining Class: org.linqs.psl.application.inference.InferenceApplication
Description: The class to use for term generator. Should be compatible with REASONER_KEY and TERM_STORE_KEY.
Key: inference.termstore
Type: String
Default Value: org.linqs.psl.reasoner.admm.term.ADMMTermStore
Module: psl-core
Defining Class: org.linqs.psl.application.inference.InferenceApplication
Description: The class to use for term storage. Should be compatible with REASONER_KEY.
Key: initialweighthyperband.internalwla
Type: String
Default Value: MaxLikelihoodMPE.class.getName()
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.InitialWeightHyperband
Description: The internal weight learning application (WLA) to use. Should actually be a VotedPerceptron.
Key: lazyatommanager.activation
Type: double
Default Value: 0.01
Module: psl-core
Defining Class: org.linqs.psl.database.atom.LazyAtomManager
Description: The minimum value an atom must take for it to be activated. Must be a flot in (0,1].
Key: lazymaxlikelihoodmpe.maxgrowrounds
Type: int
Default Value: 100
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.maxlikelihood.LazyMaxLikelihoodMPE
Description: Key for int property for the maximum number of rounds of lazy growing.
Key: lazympeinference.maxrounds
Type: int
Default Value: 100
Module: psl-core
Defining Class: org.linqs.psl.application.inference.LazyMPEInference
Description: Key for int property for the maximum number of rounds of inference.
Key: maxpiecewisepseudolikelihood.numsamples
Type: int
Default Value: 100
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.maxlikelihood.MaxPiecewisePseudoLikelihood
Description: Key for positive integer property. MaxPiecewisePseudoLikelihood will sample this many values to approximate the expectations.
Key: maxspeudolikelihood.bool
Type: boolean
Default Value: false
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.maxlikelihood.MaxPseudoLikelihood
Description: Boolean property. If true, MaxPseudoLikelihood will treat RandomVariableAtoms as boolean valued. Note that this restricts the types of contraints supported.
Key: maxspeudolikelihood.minwidth
Type: double
Default Value: 1e-2
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.maxlikelihood.MaxPseudoLikelihood
Description: Key for positive double property. Used as minimum width for bounds of integration.
Key: maxspeudolikelihood.numsamples
Type: int
Default Value: 10
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.maxlikelihood.MaxPseudoLikelihood
Description: Key for positive integer property. MaxPseudoLikelihood will sample this many values to approximate the integrals in the marginal computation.
Key: memorytermstore.initialsize
Type: int
Default Value: 5000
Module: psl-core
Defining Class: org.linqs.psl.reasoner.term.MemoryTermStore
Description: Initial size for the memory store.
Key: optimalcover.blockadvantage
Type: double
Default Value: 100.0
Module: psl-core
Defining Class: org.linqs.psl.database.rdbms.OptimalCover
Description: The cost for a blocking predicate is divided by this.
Key: optimalcover.joinadvantage
Type: double
Default Value: 2.0
Module: psl-core
Defining Class: org.linqs.psl.database.rdbms.OptimalCover
Description: The cost for a JOIN.
Key: pairedduallearner.admmsteps
Type: int
Default Value: 1
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.em.PairedDualLearner
Description: Key for Integer property that indicates how many steps of ADMM to run for each inner objective before each gradient iteration (parameter N in the ICML paper)
Key: pairedduallearner.warmuprounds
Type: int
Default Value: 0
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.em.PairedDualLearner
Description: Key for Integer property that indicates how many rounds of paired-dual learning to run before beginning to update the weights (parameter K in the ICML paper)
Key: parallel.numthreads
Type: int
Default Value: Runtime.getRuntime().availableProcessors()
Module: psl-core
Defining Class: org.linqs.psl.util.Parallel
Description:
Key: persistedatommanager.throwaccessexception
Type: boolean
Default Value: true
Module: psl-core
Defining Class: org.linqs.psl.database.atom.PersistedAtomManager
Description: Whether or not to throw an exception on illegal access. Note that in most cases, this indicates incorrectly formed data. This should only be set to true when the user understands why these exceptions are thrown in the first place and the grounding implications of not having the atom initially in the database.
Key: random.seed
Type: int
Default Value: 4
Module: psl-core
Defining Class: org.linqs.psl.util.RandUtils
Description:
Key: randomgridsearch.maxlocations
Type: int
Default Value: 150
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.grid.RandomGridSearch
Description: The max number of locations to search.
Key: rankingevaluator.representative
Type:
Default Value:
Module: psl-core
Defining Class: org.linqs.psl.evaluation.statistics.RankingEvaluator
Description: The representative metric. Default to F1. Must match a string from the RepresentativeMetric enum.
Key: rankingevaluator.threshold
Type: double
Default Value: 0.5
Module: psl-core
Defining Class: org.linqs.psl.evaluation.statistics.RankingEvaluator
Description: The truth threshold.
Key: ranksearch.scalingfactors
Type: String
Default Value: 1:2:10:100
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.search.grid.RankSearch
Description: A comma-separated list of scaling factors.
Key: rdbmsdatabase.optimalcover
Type: boolean
Default Value: false
Module: psl-core
Defining Class: org.linqs.psl.database.rdbms.RDBMSDatabase
Description: Use optimal cover grounding.
Key: votedperceptron.averagesteps
Type: boolean
Default Value: false
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: Key for Boolean property that indicates whether to average all visited weights together for final output.
Key: votedperceptron.clipnegativeweights
Type: boolean
Default Value: true
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: If true, then weight will not be allowed to go negative (clipped at zero).
Key: votedperceptron.cutobjective
Type: boolean
Default Value: false
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: If true, then cut the step size in half whenever the objective increases.
Key: votedperceptron.inertia
Type: double
Default Value: 0.00
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: The inertia that is used for adaptive step sizes. Should be in [0, 1).
Key: votedperceptron.l1regularization
Type: double
Default Value: 0.0
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: Key for positive double property scaling the L1 regularization \gamma * |w|
Key: votedperceptron.l2regularization
Type: double
Default Value: 0.0
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: Key for positive double property scaling the L2 regularization (\lambda / 2) * ||w||^2
Key: votedperceptron.numsteps
Type: int
Default Value: 25
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: Key for positive integer property. VotedPerceptron will take this many steps to learn weights.
Key: votedperceptron.scalegradient
Type: boolean
Default Value: true
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: Key for Boolean property that indicates whether to scale gradient by number of groundings
Key: votedperceptron.scalestepsize
Type: boolean
Default Value: true
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: If true, then scale the step size down by the iteration.
Key: votedperceptron.stepsize
Type: double
Default Value: 0.2
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: Key for positive double property which will be multiplied with the objective gradient to compute a step.
Key: votedperceptron.zeroinitialweights
Type: boolean
Default Value: false
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.VotedPerceptron
Description: If true, then start all weights at zero for learning.
Key: weightlearning.evaluator
Type: String
Default Value: ContinuousEvaluator.class.getName()
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.WeightLearningApplication
Description: An evalautor capable of producing a score for the current weight configuration. Child methods may use this at their own discrection. This is only used for logging/information, and not for gradients.
Key: weightlearning.groundrulestore
Type: String
Default Value: MemoryGroundRuleStore.class.getName()
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.WeightLearningApplication
Description: The class to use for ground rule storage.
Key: weightlearning.randomweights
Type: boolean
Default Value: false
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.WeightLearningApplication
Description: Randomize weights before running. The randomization will happen during ground model initialization.
Key: weightlearning.reasoner
Type: String
Default Value: ADMMReasoner.class.getName()
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.WeightLearningApplication
Description: The class to use for inference.
Key: weightlearning.termgenerator
Type: String
Default Value: ADMMTermGenerator.class.getName()
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.WeightLearningApplication
Description: The class to use for term generator. Should be compatible with REASONER_KEY and TERM_STORE_KEY.
Key: weightlearning.termstore
Type: String
Default Value: ADMMTermStore.class.getName()
Module: psl-core
Defining Class: org.linqs.psl.application.learning.weight.WeightLearningApplication
Description: The class to use for term storage. Should be compatible with REASONER_KEY.