Features

Here is a brief list of the features of particles:

• state-space models may be defined as python objects, in a basic form of probabilistic programming.
• Bootstrap filter, guided filter, auxiliary particle filter.
• exact filtering/smoothing algorithms: Kalman (linear Gaussian models), and forward-backward (finite hidden Markov models).
• Several resampling schemes are implemented.
• Sequential quasi-Monte Carlo (and related tools: Hilbert ordering, RQMC sampling).
• smoothing: on-line and off-line, O(N^2) and O(N) versions of standard algorithms (FFBS, two-filter)
• SMC samplers: IBIS (data-tempering) and SMC tempering. Static models may be defined as Python objects.
• Bayesian inference for state-space models: several PMCMC (particle MCMC algorithms are implemented), such as PMMH and Particle Gibbs. Also SMC^2.
• genealogy-based variance estimators (Chan & Lai, 2013; Lee & Whiteley, 2018; Olsson & Douc, 2019).
• a Pima indian example is included.