DiracKernel
The Dirac kernel with conditional mean prameter $\mu$ is denotd by
\[k(y\mid x) = \delta(y - \mu(x)).\]
Types
MarkovKernels.AbstractDiracKernel
— TypeAbstractDiracKernel
Abstract type for representing Dirac kernels.
MarkovKernels.DiracKernel
— TypeDiracKernel
Type for representing Dirac kernels K(y,x) = δ(y - μ(x)).
MarkovKernels.IdentityKernel
— TypeIdentityKernel
Struct for representing kernels that act like identity under marginalization.
Type aliases
const AffineDiracKernel{T} = DiracKernel{T,<:AbstractAffineMap}
Constructors
MarkovKernels.DiracKernel
— MethodDiracKernel(Φ::AbstractMatrix, Σ)
Creates a DiracKernel with a linear conditional mean function given by
x ↦ Φ * x.
MarkovKernels.DiracKernel
— MethodDiracKernel(Φ::AbstractMatrix, b::AbstractVector, Σ)
Creates a DiracKernel with an affine conditional mean function given by
x ↦ b + Φ * x.
MarkovKernels.DiracKernel
— MethodDiracKernel(Φ::AbstractMatrix, b::AbstractVector, c::AbstractVector, Σ)
Creates a DiracKernel with an affine corrector conditional mean function given by
x ↦ b + Φ * (x - c).
Basics
Statistics.mean
— Methodmean(K::AbstractDiracKernel)
Computes the conditonal mean function of the Dirac kernel K. That is, the output is callable.
Conditioning and sampling
MarkovKernels.condition
— Methodcondition(K::AbstractDiracKernel, x)
Returns a Dirac distribution corresponding to the Dirac kernel K evaluated at x.
Base.rand
— Methodrand(::AbstractRNG, K::AbstractDiracKernel, x::AbstractVector)
Computes a random vector conditionally on x with respect the the Dirac kernel K using the random number generator RNG. Equivalent to mean(K)(x).
Base.rand
— Methodrand(K::AbstractDiracKernel, x::AbstractVector)
Computes a random vector conditionally on x with respect the the Dirac kernel K using the random number generator Random.GLOBAL_RNG. Equivalent to mean(K)(x).