scalismo
API
Generated with
scalismo/scalismo.sampling.proposals/GaussianRandomWalkProposal

GaussianRandomWalkProposal

scalismo.sampling.proposals.GaussianRandomWalkProposal
See theGaussianRandomWalkProposal companion object
class GaussianRandomWalkProposal(stddev: Double, val tag: String)(implicit rng: Random) extends MHProposalGenerator[DenseVector[Double]]

Classical Random Walk proposal, where the current state is perturbed using a step, which is generated by an isotropic gaussian with the given standard deviation.

Attributes

Companion:
object
Graph
Supertypes
class MHProposalGenerator[DenseVector[Double]]
trait TransitionProbability[MHSample[DenseVector[Double]]]
trait TransitionRatio[MHSample[DenseVector[Double]]]
trait ProposalGenerator[MHSample[DenseVector[Double]]]
class Object
trait Matchable
class Any
Self type

Members list

Concise view

Value members

Concrete methods

override def logTransitionProbability(from: MHSample[DenseVector[Double]], to: MHSample[DenseVector[Double]]): Double

rate of transition from to (log value)

rate of transition from to (log value)

Attributes

Definition Classes
def partial(range: Range): GaussianRandomWalkProposal

Create a new GaussianRandomWalkProposal, which only updates the coefficients in an given range.

Create a new GaussianRandomWalkProposal, which only updates the coefficients in an given range.

Attributes

override def propose(sample: MHSample[DenseVector[Double]]): MHSample[DenseVector[Double]]

draw a sample from this proposal distribution, may depend on current state

draw a sample from this proposal distribution, may depend on current state

Attributes

Definition Classes

Inherited methods

def forType[T](implicit conversion: ParameterConversion[DenseVector[Double], T]): MHProposalGenerator[T]

Applies a parameter conversion of the sampled type. With this mechanism it is, for example possible to derive a generator for a Type T (e.g. ShapeParameters) from a generator that samples only DenseVectors.

Applies a parameter conversion of the sampled type. With this mechanism it is, for example possible to derive a generator for a Type T (e.g. ShapeParameters) from a generator that samples only DenseVectors.

Attributes

Inherited from:
MHProposalGenerator
override def logTransitionRatio(from: MHSample[DenseVector[Double]], to: MHSample[DenseVector[Double]]): Double

transition ratio forward probability / backward probability

transition ratio forward probability / backward probability

Attributes

Definition Classes
Inherited from:
TransitionProbability
def relabel(generatedBy: String): MHProposalGenerator[A]

returns a new ProposalGenerator, with a new label. This is useful to summarize complicated labels in composed proposal generators

returns a new ProposalGenerator, with a new label. This is useful to summarize complicated labels in composed proposal generators

Attributes

Inherited from:
MHProposalGenerator

Concrete fields

val tag: String

Implicits

Implicits

implicit val randBasis: RandBasis
Generated with