pypomp.PanelPomp.pfilter

PanelPomp.pfilter(J: int, key: Array | None = None, theta: PanelParameters | None = None, thresh: float = 0.0, reps: int = 1, chunk_size: int = 1, CLL: bool = False, ESS: bool = False, filter_mean: bool = False, prediction_mean: bool = False) → None

Run the bootstrap particle filter (SMC) algorithm.

Evaluates the likelihood of the panel data at the specified parameter values.

Parameters:

  • J (int) – Number of particles per unit.

  • key (jax.Array or None, optional) – JAX random key. If None, uses the model’s fresh_key.

  • theta (PanelParameters or None, optional) – Parameter sets to use. If None, defaults to self.theta.

  • thresh (float, optional) – Resampling threshold. If 0.0 (default), always resample at each observation time.

  • reps (int, optional) – Number of replicates per parameter set. Defaults to 1.

  • chunk_size (int, optional) – Number of units to process per batch. Defaults to 1.

  • CLL (bool, optional) – Whether to compute conditional log-likelihoods. Defaults to False.

  • ESS (bool, optional) – Whether to compute effective sample sizes. Defaults to False.

  • filter_mean (bool, optional) – Whether to compute filtering state means. Defaults to False.

  • prediction_mean (bool, optional) – Whether to compute predicted state means. Defaults to False.

Returns:

Updates the unit-specific log-likelihoods self.theta.logLik_unit and appends a Result to the history.

Return type:

None