TimeSeriesPosterior

class probnum.filtsmooth.TimeSeriesPosterior(locations=None, states=None)

Bases: ABC

Posterior Distribution over States after time-series algorithms such as filtering/smoothing or solving ODEs.

Parameters:
  • locations (Optional[Iterable[FloatLike]]) – Locations of the posterior states (represented as random variables).

  • states (Optional[Iterable[randvars.RandomVariable]]) – Posterior random variables.

Attributes Summary

frozen

Whether the posterior is frozen.

locations

Locations of the states of the posterior.

states

States of the posterior.

Methods Summary

__call__(t)

Evaluate the time-continuous posterior at location t

append(location, state)

Append a state to the posterior.

freeze()

Freeze the posterior.

interpolate(t[, previous_index, next_index])

Evaluate the posterior at a measurement-free point.

sample(rng[, t, size])

Draw samples from the filtering/smoothing posterior.

transform_base_measure_realizations(...)

Transform base-measure-realizations into posteriors samples.

Attributes Documentation

frozen

Whether the posterior is frozen.

locations

Locations of the states of the posterior.

states

States of the posterior.

Methods Documentation

__call__(t)[source]

Evaluate the time-continuous posterior at location t

Algorithm: 1. Find closest t_prev and t_next, with t_prev < t < t_next 2. Predict from t_prev to t 3. (if self._with_smoothing=True) Predict from t to t_next 4. (if self._with_smoothing=True) Smooth from t_next to t 5. Return random variable for time t

Parameters:

t (ArrayLike) – Location, or time, at which to evaluate the posterior.

Raises:

ValueError – If time-points are not strictly increasing.

Returns:

Estimate of the states at time t.

Return type:

randvars.RandomVariable or randvars._RandomVariableList

append(location, state)[source]

Append a state to the posterior.

Parameters:
Return type:

None

freeze()[source]

Freeze the posterior.

Return type:

None

abstract interpolate(t, previous_index=None, next_index=None)[source]

Evaluate the posterior at a measurement-free point.

Returns:

Dense evaluation.

Return type:

randvars.RandomVariable or randvars._RandomVariableList

Parameters:
abstract sample(rng, t=None, size=())[source]

Draw samples from the filtering/smoothing posterior.

If nothing is specified, a single sample is drawn (supported on self.locations). If locations are specified, a single sample is drawn on those locations. If size is specified, more than a single sample is drawn.

Internally, samples from a base measure are drawn and transformed via self.transform_base_measure_realizations.

Parameters:
  • rng (Generator) – Random number generator.

  • t (ArrayLike | None) – Locations on which the samples are wanted. Default is none, which implies that self.location is used.

  • size (ShapeLike | None) – Indicates how many samples are drawn. Default is an empty tuple, in which case a single sample is returned.

Returns:

Drawn samples. If size has shape (A1, …, Z1), locations have shape (L,), and the state space model has shape (A2, …, Z2), the output has shape (A1, …, Z1, L, A2, …, Z2). For example: size=4, len(locations)=4, dim=3 gives shape (4, 4, 3).

Return type:

np.ndarray

abstract transform_base_measure_realizations(base_measure_realizations, t)[source]

Transform base-measure-realizations into posteriors samples.

Parameters:
  • base_measure_realizations (ndarray) – Base measure realizations.

  • t (ArrayLike | None) – Locations on which the transformed realizations shall represent realizations from the posterior.

Returns:

Transformed realizations.

Return type:

np.ndarray