Inverse Problem

See also

See Terminology & Notation for definitions of prior, posterior, forward operator, and other inverse problem terminology.

Constructor

InverseProblem(obs, prior, forward_operator, ...)

Inverse problem combining observations, priors, and forward model.

Inputs

InverseProblem.obs	Observation vector.
InverseProblem.prior	Prior state vector.
InverseProblem.forward_operator	Forward operator mapping state space to observation space.
InverseProblem.prior_error	Covariance matrix representing prior error.
InverseProblem.modeldata_mismatch	Covariance matrix representing model-data mismatch (observation error).
InverseProblem.constant	Optional constant term added to the forward model (e.g., background or bias).

Solving the Inverse Problem

InverseProblem.estimator	Return the fitted estimator (raises if not solved).
InverseProblem.solve(estimator, **kwargs)	Solve the inverse problem using the specified estimator.

Results

InverseProblem.posterior	Posterior state estimate.
InverseProblem.posterior_error	Posterior error covariance.
InverseProblem.prior_obs	Modeled observations using the prior (H @ prior).
InverseProblem.posterior_obs	Modeled observations (H @ posterior).
InverseProblem.kalman_gain	Kalman gain matrix (K).
InverseProblem.averaging_kernel	Averaging kernel matrix (A).

Selection and indexing

InverseProblem.get_block(component, block[, ...])

Get block from a component (Vector or Matrix).

Serialization

InverseProblem.from_file(path)	Load object from a pickle file.
InverseProblem.to_file(path)	Save object to a pickle file.