matflow.subset_simulation.subset_simulation_result.LevelSamplingResult#

class matflow.subset_simulation.subset_simulation_result.LevelSamplingResult(chains, debug_data=None)#

Result of generating one conditional Subset Simulation level.

Methods

Attributes

`component_acceptance_rate`	Return the aggregate MMH component acceptance rate.
`debug_data`	Optional algorithm-specific debugging information.
`g`	Fine performance values associated with the sampled states.
`jump_distance_sum`	Sum of RMS jump distances over all outer-chain transitions.
`jump_distances`	Return per-transition RMS jump distances, if retained.
`mean_jump_distance`	Return the mean RMS distance between adjacent stored states.
`num_chains`	Number of Markov chains used to generate the level.
`num_coarse_evals`	Total number of coarse performance-function evaluations.
`num_components_accepted`	Total number of MMH component proposals accepted.
`num_components_proposed`	Total number of MMH component proposals attempted.
`num_fine_evals`	Total number of fine performance-function evaluations.
`num_outer_moves`	Total number of stored outer-chain transitions that change state.
`num_states`	Total number of stored states across all chains.
`num_subset_accepts`	Total number of candidates passing the fine subset condition.
`num_subset_trials`	Total number of candidates tested against the fine subset condition.
`num_transitions`	Total number of attempted outer-chain transitions.
`outer_move_rate`	Return the fraction of stored outer-chain transitions that move.
`subset_acceptance_rate`	Return the fraction of candidates passing the fine subset condition.
`x`	Sampled states, arranged as `(chains, states, dimensions)`.
`chains`	Results for the individual Markov chains used to generate this level.

Parameters:

chains: tuple[MarkovChainResult, ...]#: Results for the individual Markov chains used to generate this level.

property component_acceptance_rate: float#

Return the aggregate MMH component acceptance rate.

The rate includes MMH component proposals belonging to candidate states that subsequently fail the Subset Simulation condition.

debug_data: dict[str, Any] | None = None#: Optional algorithm-specific debugging information.

property g: ndarray#: Fine performance values associated with the sampled states.

property jump_distance_sum: float#: Sum of RMS jump distances over all outer-chain transitions.

property jump_distances: ndarray | None#

Return per-transition RMS jump distances, if retained.

The resulting array has shape (chains, transitions). The value is None if detailed jump distances were not retained for every chain.

property mean_jump_distance: float#

Return the mean RMS distance between adjacent stored states.

Rejected outer transitions contribute a zero jump distance.

property num_chains: int#: Number of Markov chains used to generate the level.

property num_coarse_evals: int#: Total number of coarse performance-function evaluations.

property num_components_accepted: int#: Total number of MMH component proposals accepted.

property num_components_proposed: int#: Total number of MMH component proposals attempted.

property num_fine_evals: int#: Total number of fine performance-function evaluations.

property num_outer_moves: int#: Total number of stored outer-chain transitions that change state.

property num_states: int#: Total number of stored states across all chains.

property num_subset_accepts: int#: Total number of candidates passing the fine subset condition.

property num_subset_trials: int#: Total number of candidates tested against the fine subset condition.

property num_transitions: int#: Total number of attempted outer-chain transitions.

property outer_move_rate: float#: Return the fraction of stored outer-chain transitions that move.

property subset_acceptance_rate: float#: Return the fraction of candidates passing the fine subset condition.

property x: ndarray#: Sampled states, arranged as (chains, states, dimensions).