pypomp.PanelPomp.probe

PanelPomp.probe(probes: dict[str, Callable[[DataFrame], float]], key: Array, nsim: int = 100, theta: PanelParameters | None = None) → DataFrame

Evaluate probe statistics on real and simulated data for each unit.

Parameters:

  • probes (dict of str to callable) – Dictionary mapping probe names to functions. Each function must take a dataframe of observations for a single unit and return a scalar float.

  • key (jax.Array) – JAX random key.

  • nsim (int, optional) – Number of simulations to run per replicate. Defaults to 100.

  • theta (PanelParameters or None, optional) – Parameters to simulate from. If None, defaults to self.theta.

Returns:

Long-format DataFrame with columns:

probe, value, is_real_data, theta_idx, sim, unit

Return type:

pd.DataFrame