cpf.sisr: Conditional Particle filter via Sequential Importance...

Description Usage Arguments Details Value

Description

Runs a conditional particle filter on a given Non-Linear State Space models

Usage

1
2
3
4
5
  cpf.sisr(x, nlss, y, N, proposal.rnd = prior.rnd,
    proposal.logpdf = prior.logpdf,
    resampling = MultinomialR,
    initial.proposal.rnd = initial.rnd,
    initial.proposal.logpdf = initial.logpdf)

Arguments

x

Trajectory on which to condition

...

same as classical sisr

Details

Warning: the resampling is Multinomial. Residual will bias the sampler, due to the way we overried the Nth particle by the conditioned trajectory.

Value

A list with the same components as SISR, plus the components:

x

Array (T, D) of the new selected trajectory


nickpoison/nltsa documentation built on May 23, 2019, 4:48 p.m.