amici.petab_objective

PEtab Objective

Functionality related to running simulations or evaluating the objective function as defined by a PEtab problem

Functions

aggregate_sllh(amici_model, rdatas, ...[, ...])	Aggregate likelihood gradient for all conditions, according to PEtab parameter mapping.
create_edata_for_condition(condition, ...)	Get amici.amici.ExpData for the given PEtab condition.
create_edatas(amici_model, petab_problem[, ...])	Create list of amici.amici.ExpData objects for PEtab problem.
create_parameter_mapping(petab_problem, ...)	Generate AMICI specific parameter mapping.
create_parameter_mapping_for_condition(...)	Generate AMICI specific parameter mapping for condition.
create_parameterized_edatas(amici_model, ...)	Create list of :class:amici.ExpData objects with parameters filled in.
rdatas_to_measurement_df(rdatas, model, ...)	Create a measurement dataframe in the PEtab format from the passed rdatas and own information.
rdatas_to_simulation_df(rdatas, model, ...)	Create a PEtab simulation dataframe from amici.amici.ReturnData s.
rescale_sensitivity(sensitivity, ...)	Rescale a sensitivity between parameter scales.
simulate_petab(petab_problem, amici_model[, ...])	Simulate PEtab model.

amici.petab_objective.aggregate_sllh(amici_model, rdatas, parameter_mapping, edatas, petab_scale=True, petab_problem=None)

Aggregate likelihood gradient for all conditions, according to PEtab parameter mapping.

Parameters

• amici_model (typing.Union[amici.amici.Model, amici.amici.ModelPtr]) – AMICI model from which rdatas were obtained.

• rdatas (typing.Sequence[amici.numpy.ReturnDataView]) – Simulation results.

• parameter_mapping (typing.Optional[amici.parameter_mapping.ParameterMapping]) – PEtab parameter mapping to condition-specific simulation parameters.

• edatas (typing.List[typing.Union[amici.swig_wrappers.ExpData, amici.amici.ExpDataPtr]]) – Experimental data used for simulation.

• petab_scale (bool) – Whether to check that sensitivities were computed with parameters on the scales provided in the PEtab parameters table.

• petab_problem (typing.Optional[petab.problem.Problem]) – The PEtab problem that defines the parameter scales.

Return type

typing.Optional[typing.Dict[str, float]]

Returns

Aggregated likelihood sensitivities.

amici.petab_objective.create_edata_for_condition(condition, measurement_df, amici_model, petab_problem, observable_ids)

Get amici.amici.ExpData for the given PEtab condition.

Sets timepoints, observed data and sigmas.

Parameters

• condition (typing.Union[typing.Dict, pandas.core.series.Series]) – pandas.DataFrame row with preequilibrationConditionId and simulationConditionId.

• measurement_df (pandas.core.frame.DataFrame) – pandas.DataFrame with measurements for the given condition.

• amici_model (typing.Union[amici.amici.Model, amici.amici.ModelPtr]) – AMICI model

• petab_problem (petab.problem.Problem) – Underlying PEtab problem

• observable_ids (typing.List[str]) – List of observable IDs

Return type

amici.swig_wrappers.ExpData

Returns

ExpData instance.

amici.petab_objective.create_edatas(amici_model, petab_problem, simulation_conditions=None)

Create list of amici.amici.ExpData objects for PEtab problem.

Parameters

• amici_model (typing.Union[amici.amici.Model, amici.amici.ModelPtr]) – AMICI model.

• petab_problem (petab.problem.Problem) – Underlying PEtab problem.

• simulation_conditions (typing.Union[pandas.core.frame.DataFrame, typing.Dict, None]) – Result of petab.get_simulation_conditions(). Can be provided to save time if this has be obtained before.

Return type

typing.List[amici.swig_wrappers.ExpData]

Returns

List with one amici.amici.ExpData per simulation condition, with filled in timepoints and data.

amici.petab_objective.create_parameter_mapping(petab_problem, simulation_conditions, scaled_parameters, amici_model, **parameter_mapping_kwargs)

Generate AMICI specific parameter mapping.

Parameters

• petab_problem (petab.problem.Problem) – PEtab problem

• simulation_conditions (typing.Union[pandas.core.frame.DataFrame, typing.List[typing.Dict]]) – Result of petab.get_simulation_conditions(). Can be provided to save time if this has been obtained before.

• scaled_parameters (bool) – If True, problem_parameters are assumed to be on the scale provided in the PEtab parameter table and will be unscaled. If False, they are assumed to be in linear scale.

• amici_model (typing.Union[amici.amici.Model, amici.amici.ModelPtr]) – AMICI model.

• parameter_mapping_kwargs – Optional keyword arguments passed to petab.get_optimization_to_simulation_parameter_mapping(). To allow changing fixed PEtab problem parameters (estimate=0), use fill_fixed_parameters=False.

Return type

amici.parameter_mapping.ParameterMapping

Returns

List of the parameter mappings.

amici.petab_objective.create_parameter_mapping_for_condition(parameter_mapping_for_condition, condition, petab_problem, amici_model)

Generate AMICI specific parameter mapping for condition.

Parameters

• parameter_mapping_for_condition (typing.Tuple[typing.Dict[str, typing.Union[str, numbers.Number]], typing.Dict[str, typing.Union[str, numbers.Number]], typing.Dict[str, str], typing.Dict[str, str]]) – Preliminary parameter mapping for condition.

• condition (typing.Union[pandas.core.series.Series, typing.Dict]) – pandas.DataFrame row with preequilibrationConditionId and simulationConditionId.

• petab_problem (petab.problem.Problem) – Underlying PEtab problem.

• amici_model (typing.Union[amici.amici.Model, amici.amici.ModelPtr]) – AMICI model.

Return type

amici.parameter_mapping.ParameterMappingForCondition

Returns

The parameter and parameter scale mappings, for fixed preequilibration, fixed simulation, and variable simulation parameters, and then the respective scalings.

amici.petab_objective.create_parameterized_edatas(amici_model, petab_problem, problem_parameters, scaled_parameters=False, parameter_mapping=None, simulation_conditions=None)

Create list of :class:amici.ExpData objects with parameters filled in.

Parameters

• amici_model (typing.Union[amici.amici.Model, amici.amici.ModelPtr]) – AMICI Model assumed to be compatible with petab_problem.

• petab_problem (petab.problem.Problem) – PEtab problem to work on.

• problem_parameters (typing.Dict[str, numbers.Number]) – Run simulation with these parameters. If None, PEtab nominalValues will be used. To be provided as dict, mapping PEtab problem parameters to SBML IDs.

• scaled_parameters (bool) – If True, problem_parameters are assumed to be on the scale provided in the PEtab parameter table and will be unscaled. If False, they are assumed to be in linear scale.

• parameter_mapping (typing.Optional[amici.parameter_mapping.ParameterMapping]) – Optional precomputed PEtab parameter mapping for efficiency, as generated by create_parameter_mapping().

• simulation_conditions (typing.Union[pandas.core.frame.DataFrame, typing.Dict, None]) – Result of petab.get_simulation_conditions(). Can be provided to save time if this has been obtained before.

Return type

typing.List[amici.swig_wrappers.ExpData]

Returns

List with one amici.amici.ExpData per simulation condition, with filled in timepoints, data and parameters.

amici.petab_objective.rdatas_to_measurement_df(rdatas, model, measurement_df)

Create a measurement dataframe in the PEtab format from the passed rdatas and own information.

Parameters

• rdatas (typing.Sequence[amici.amici.ReturnData]) – A sequence of rdatas with the ordering of petab.get_simulation_conditions().

• model (typing.Union[amici.amici.Model, amici.amici.ModelPtr]) – AMICI model used to generate rdatas.

• measurement_df (pandas.core.frame.DataFrame) – PEtab measurement table used to generate rdatas.

Return type

pandas.core.frame.DataFrame

Returns

A dataframe built from the rdatas in the format of measurement_df.

amici.petab_objective.rdatas_to_simulation_df(rdatas, model, measurement_df)

Create a PEtab simulation dataframe from amici.amici.ReturnData s.

See rdatas_to_measurement_df() for details, only that model outputs will appear in column simulation instead of measurement.

Return type

pandas.core.frame.DataFrame

amici.petab_objective.rescale_sensitivity(sensitivity, parameter_value, old_scale, new_scale)

Rescale a sensitivity between parameter scales.

Parameters

• sensitivity (float) – The sensitivity corresponding to the parameter value.

• parameter_value (float) – The parameter vector element, on old_scale.

• old_scale (str) – The scale of the parameter value.

• new_scale (str) – The parameter scale on which to rescale the sensitivity.

Return type

float

Returns

The rescaled sensitivity.

amici.petab_objective.simulate_petab(petab_problem, amici_model, solver=None, problem_parameters=None, simulation_conditions=None, edatas=None, parameter_mapping=None, scaled_parameters=False, log_level=30, num_threads=1, failfast=True, scaled_gradients=False)

Simulate PEtab model.

Note

Regardless of scaled_parameters, unscaled sensitivities are returned, unless scaled_gradients=True.

Parameters

• petab_problem (petab.problem.Problem) – PEtab problem to work on.

• amici_model (typing.Union[amici.amici.Model, amici.amici.ModelPtr]) – AMICI Model assumed to be compatible with petab_problem.

• solver (typing.Optional[amici.amici.Solver]) – An AMICI solver. Will use default options if None.

• problem_parameters (typing.Optional[typing.Dict[str, float]]) – Run simulation with these parameters. If None, PEtab nominalValues will be used. To be provided as dict, mapping PEtab problem parameters to SBML IDs.

• simulation_conditions (typing.Union[pandas.core.frame.DataFrame, typing.Dict, None]) – Result of petab.get_simulation_conditions(). Can be provided to save time if this has be obtained before. Not required if edatas and parameter_mapping are provided.

• edatas (typing.Optional[typing.List[typing.Union[amici.swig_wrappers.ExpData, amici.amici.ExpDataPtr]]]) – Experimental data. Parameters are inserted in-place for simulation.

• parameter_mapping (typing.Optional[amici.parameter_mapping.ParameterMapping]) – Optional precomputed PEtab parameter mapping for efficiency, as generated by create_parameter_mapping().

• scaled_parameters (typing.Optional[bool]) – If True, problem_parameters are assumed to be on the scale provided in the PEtab parameter table and will be unscaled. If False, they are assumed to be in linear scale.

• log_level (int) – Log level, see amici.logging module.

• num_threads (int) – Number of threads to use for simulating multiple conditions (only used if compiled with OpenMP).

• failfast (bool) – Returns as soon as an integration failure is encountered, skipping any remaining simulations.

• scaled_gradients (bool) – Whether to compute gradients on parameter scale (True) or not (False).

Return type

typing.Dict[str, typing.Any]

Returns

Dictionary of

• cost function value (LLH),

• list of amici.amici.ReturnData (RDATAS),

• list of amici.amici.ExpData (EDATAS),

corresponding to the different simulation conditions. For ordering of simulation conditions, see petab.Problem.get_simulation_conditions_from_measurement_df().