OutputConfig

Bases: CUDAFactoryConfig

Validated configuration for solver outputs and summaries.

Parameters:

max_states – Maximum number of state variables.
max_observables – Maximum number of observable variables.
saved_state_indices – Indices of state variables to save. Defaults to an empty collection that resolves to all states.
saved_observable_indices – Indices of observable variables to save. Defaults to an empty collection that resolves to all observables.
summarised_state_indices – Indices of state variables to summarise. Defaults to saved_state_indices.
summarised_observable_indices – Indices of observable variables to summarise. Defaults to saved_observable_indices.
output_types – Requested output type names, including summary metric identifiers.
sample_summaries_every – Time interval between summary metric samples. Used by derivative metrics to scale finite differences. Defaults to 0.01 seconds.
precision (type[numpy.float16] | type[numpy.float32] | type[numpy.float64] | numpy.dtype[numpy.float16] | numpy.dtype[numpy.float32] | numpy.dtype[numpy.float64]) – Numerical precision for output calculations. Defaults to np.float32.

Notes

Private attributes store numpy arrays so that properties can manage circular dependencies between index validation and flag updates. The post-initialisation hook applies default indices, validates bounds, and ensures at least one output path is active.

_check_for_no_outputs() → None[source]

Ensure that at least one output type is requested.

Returns:: Returns None.
Return type:: None
Raises:: ValueError – Raised when no output types are enabled.

_check_saved_indices() → None[source]

Convert saved indices to numpy arrays.

Returns:: Returns None.
Return type:: None

Notes

Converts index collections to numpy int arrays.

_check_summarised_indices() → None[source]

Convert summarised indices to numpy arrays.

Returns:: Returns None.
Return type:: None

Notes

Converts index collections to numpy int arrays.

_validate_index_arrays() → None[source]

Validate all index arrays for bounds and duplication.

Returns:: Returns None.
Return type:: None

Notes

Called post-initialisation so that None values can be replaced by defaults before validation.

property buffer_sizes_dict: dict[str, int]: Returns a dict of buffer sizes to update other objects’ settings

property compile_flags: OutputCompileFlags

Get compile flags for this configuration.

Returns:: Flags indicating which output features should be compiled.
Return type:: OutputCompileFlags

Create configuration from integrator-compatible specifications.

Parameters:

output_types – Output types including "state", "observables", "time", and summary metric names such as "max" or "rms".
saved_state_indices – Indices of states to save in time-domain output.
saved_observable_indices – Indices of observables to save in time-domain output.
summarised_state_indices – Indices of states for summary calculations. Defaults to saved_state_indices.
summarised_observable_indices – Indices of observables for summary calculations. Defaults to saved_observable_indices.
max_states – Total number of state variables in the system.
max_observables – Total number of observable variables in the system.
sample_summaries_every – Time interval between summary metric samples. Used by derivative metrics to scale finite differences. Defaults to 0.01.
precision – Numerical precision for output calculations. Defaults to np.float32 if not provided.

Returns:

Configured output configuration object.

Return type:

OutputConfig

Notes

This class method provides a convenient interface for creating OutputConfig objects from the parameter format used by integrator classes. It handles None values appropriately by converting them to empty arrays.

property max_observables: int: Maximum number of observables.

property max_states: int: Maximum number of states.

property n_saved_observables: int

Get number of observables that will be saved.

Returns:: Number of observable variables to save in time-domain output.
Return type:: int

property n_saved_states: int

Get number of states that will be saved.

Returns:: Number of state variables to save in time-domain output.
Return type:: int

Notes

Returns the length of saved_state_indices when save_state is True, otherwise 0.

property n_summarised_observables: int

Get number of observables that will be summarised.

Returns:: Number of observable variables for summary calculations.
Return type:: int

property n_summarised_states: int

Get number of states that will be summarised.

Returns:: Number of state variables for summary calculations.
Return type:: int

Notes

Returns the length of summarised_state_indices when save_summaries is active, otherwise 0.

property observable_summaries_buffer_height: int

Calculate total buffer height for observable summaries.

Returns:: Total buffer height for all observable summary calculations.
Return type:: int

property observable_summaries_output_height: int

Calculate total output height for observable summaries.

Returns:: Total output height for all observable summary results.
Return type:: int

property output_types: List[str]: Configured output type names in declaration order.

property sample_summaries_every: float | None: Time interval between summary metric samples.

property save_counters: bool: Whether iteration counters should be saved.

property save_observables: bool: Whether observable saving is enabled with valid indices.

property save_state: bool: Whether state saving is enabled with valid indices.

property save_summaries: bool: Whether any summary metric is configured.

property save_time: bool: Whether solver time samples should be saved.

property saved_observable_indices: ndarray[Any, dtype[int64]]: Observable indices to save, or an empty array when disabled.

property saved_state_indices: ndarray[Any, dtype[int64]]: State indices to save, or an empty array when disabled.

property state_summaries_buffer_height: int

Calculate total buffer height for state summaries.

Returns:: Total buffer height for all state summary calculations.
Return type:: int

property state_summaries_output_height: int

Calculate total output height for state summaries.

Returns:: Total output height for all state summary results.
Return type:: int

property summaries_buffer_height_per_var: int

Calculate buffer size per variable for summary calculations.

Returns:: Buffer height required per variable for summary metrics.
Return type:: int

property summaries_output_height_per_var: int

Calculate output array height per variable for summaries.

Returns:: Output height required per variable for summary results.
Return type:: int

property summarise_observables: bool: Whether observable variables contribute to summaries.

property summarise_state: bool: Whether state variables contribute to summaries.

property summarised_observable_indices: ndarray[Any, dtype[int64]]: Observable indices for summaries, or an empty array when disabled.

property summarised_state_indices: ndarray[Any, dtype[int64]]: State indices for summaries, or an empty array when disabled.

property summary_legend_per_variable: dict[int, str]

Map per-variable summary indices to metric names.

Returns:: Dictionary that assigns each per-variable summary slot to its metric identifier.
Return type:: dict[int, str]

property summary_parameters: dict[str, object]

Collect parameters required by the registered summary metrics.

Returns:: Dictionary of metric-specific keyword arguments needed during compilation.
Return type:: dict[str, object]

property summary_types: Tuple[str, ...]: Configured summary metric identifiers.

property summary_unit_modifications: dict[int, str]

Map per-variable summary indices to unit modifications.

Returns:: Dictionary that assigns each per-variable summary slot to its unit modification string.
Return type:: dict[int, str]

property total_summary_buffer_size: int

Calculate total size of all summary buffers.

Returns:: Combined size of state and observable summary buffers.
Return type:: int

update_from_outputs_list(output_types: list[str]) → None[source]

Update configuration from a list of output type names.

Parameters:: output_types – Output type names to configure.
Returns:: Returns None.
Return type:: None

Notes

Parses the list to set boolean flags and collect summary metric selections. Unknown entries trigger warnings but do not raise exceptions.

validation_passes() → bool[source]