amici.amici

Core C++ bindings

This module encompasses the complete public C++ API of AMICI, which was exposed via swig. All functions listed here are directly accessible in the main amici package, i.e., amici.amici.ExpData is available as amici.ExpData. Usage of functions and classes from the base amici package is generally recommended as they often include convenience wrappers that avoid common pitfalls when accessing C++ types from python and implement some nonstandard type conversions.

Classes

BoolVector(*args)	Swig-Generated class templating common python types including Iterable [bool] and numpy.array [bool] to facilitate interfacing with C++ bindings.
ContextManager()	Generic implementation for a context manager, explicitly deletes copy and move operators for derived classes
DoubleVector(*args)	Swig-Generated class templating common python types including Iterable [float] and numpy.array [float] to facilitate interfacing with C++ bindings.
ExpData(*args)	ExpData carries all information about experimental or condition-specific data
ExpDataPtr(*args)	Swig-Generated class that implements smart pointers to ExpData as objects.
ExpDataPtrVector(*args)	Swig-Generated class templating common python types including Iterable [amici.amici.ExpData] and numpy.array [amici.amici.ExpData] to facilitate interfacing with C++ bindings.
FixedParameterContext(value)	An enumeration.
IntVector(*args)	Swig-Generated class templating common python types including Iterable [int] and numpy.array [int] to facilitate interfacing with C++ bindings.
InternalSensitivityMethod(value)	An enumeration.
InterpolationType(value)	An enumeration.
LinearMultistepMethod(value)	An enumeration.
LinearSolver(value)	An enumeration.
Model(*args, **kwargs)	The Model class represents an AMICI ODE/DAE model.
ModelContext(model)	The ModelContext temporarily stores amici::Model::state and restores it when going out of scope
ModelPtr(*args)	Swig-Generated class that implements smart pointers to Model as objects.
NewtonDampingFactorMode(value)	An enumeration.
NonlinearSolverIteration(value)	An enumeration.
ParameterScaling(value)	An enumeration.
ParameterScalingVector(*args)	Swig-Generated class, which, in contrast to other Vector classes, does not allow for simple interoperability with common python types, but must be created using amici.amici.parameterScalingFromIntVector()
RDataReporting(value)	An enumeration.
ReturnData(*args)	Stores all data to be returned by amici.amici.runAmiciSimulation().
ReturnDataPtr(*args)	Swig-Generated class that implements smart pointers to ReturnData as objects.
SecondOrderMode(value)	An enumeration.
SensitivityMethod(value)	An enumeration.
SensitivityOrder(value)	An enumeration.
Solver(*args, **kwargs)	The Solver class provides a generic interface to CVODES and IDAS solvers, individual realizations are realized in the CVodeSolver and the IDASolver class.
SolverPtr(*args)	Swig-Generated class that implements smart pointers to Solver as objects.
SteadyStateSensitivityMode(value)	An enumeration.
SteadyStateStatus(value)	An enumeration.
SteadyStateStatusVector(*args)
StringDoubleMap(*args)	Swig-Generated class templating Dict [str, float] to facilitate interfacing with C++ bindings.
StringVector(*args)	Swig-Generated class templating common python types including Iterable [str] and numpy.array [str] to facilitate interfacing with C++ bindings.

Functions Summary

backtraceString(maxFrames)	Returns the current backtrace as std::string
compiledWithOpenMP()	rtype bool
enum(prefix)
getScaledParameter(unscaledParameter, scaling)	Apply parameter scaling according to scaling
getUnscaledParameter(scaledParameter, scaling)	Remove parameter scaling according to scaling
parameterScalingFromIntVector(intVec)	rtype amici.amici.ParameterScalingVector
runAmiciSimulation(solver, edata, model[, …])	Core integration routine.
runAmiciSimulations(solver, edatas, model, …)	Same as runAmiciSimulation, but for multiple ExpData instances.
scaleParameters(bufferUnscaled, pscale, …)	Apply parameter scaling according to scaling
unscaleParameters(bufferScaled, pscale, …)	Remove parameter scaling according to the parameter scaling in pscale

Functions

amici.amici.backtraceString(maxFrames: int) → str

Returns the current backtrace as std::string

Parameters

maxFrames (int) – Number of frames to include

Return type

string

Returns

Backtrace

amici.amici.compiledWithOpenMP() → bool

Return type

bool

amici.amici.enum(prefix)

amici.amici.getScaledParameter(unscaledParameter: float, scaling: amici.amici.ParameterScaling) → float

Apply parameter scaling according to scaling

Parameters

• unscaledParameter (float) –

• scaling (amici.amici.ParameterScaling) – parameter scaling

Return type

float

Returns

Scaled parameter

amici.amici.getUnscaledParameter(scaledParameter: float, scaling: amici.amici.ParameterScaling) → float

Remove parameter scaling according to scaling

Parameters

• scaledParameter (float) – scaled parameter

• scaling (amici.amici.ParameterScaling) – parameter scaling

Return type

float

Returns

Unscaled parameter

amici.amici.parameterScalingFromIntVector(intVec: amici.amici.IntVector) → amici.amici.ParameterScalingVector

Return type

amici.amici.ParameterScalingVector

amici.amici.runAmiciSimulation(solver: amici.amici.Solver, edata: amici.amici.ExpData, model: amici.amici.Model, rethrow: bool = False) → amici.amici.ReturnData

Core integration routine. Initializes the solver and runs the forward and backward problem.

Parameters

• solver (amici.amici.Solver) – Solver instance

• edata (amici.amici.ExpData) – pointer to experimental data object

• model (amici.amici.Model) – model specification object

• rethrow (bool) – rethrow integration exceptions?

Return type

ReturnData

Returns

rdata pointer to return data object

amici.amici.runAmiciSimulations(solver: amici.amici.Solver, edatas: amici.amici.ExpDataPtrVector, model: amici.amici.Model, failfast: bool, num_threads: int) → Iterable[amici.amici.ReturnData]

Same as runAmiciSimulation, but for multiple ExpData instances. When compiled with OpenMP support, this function runs multi-threaded.

Parameters

• solver (amici.amici.Solver) – Solver instance

• edatas (amici.amici.ExpDataPtrVector) – experimental data objects

• model (amici.amici.Model) – model specification object

• failfast (bool) – flag to allow early termination

• num_threads (int) – number of threads for parallel execution

Return type

Iterable[ReturnData]

Returns

vector of pointers to return data objects

amici.amici.scaleParameters(bufferUnscaled: Iterable[float], pscale: Iterable[amici.amici.ParameterScaling], bufferScaled: Iterable[float]) → None

Apply parameter scaling according to scaling

Parameters

• bufferUnscaled (typing.Iterable[float]) –

• pscale (typing.Iterable[amici.amici.ParameterScaling]) – parameter scaling

• bufferScaled (typing.Iterable[float]) – destination

Return type

None

amici.amici.unscaleParameters(bufferScaled: Iterable[float], pscale: Iterable[amici.amici.ParameterScaling], bufferUnscaled: Iterable[float]) → None

Remove parameter scaling according to the parameter scaling in pscale

All vectors must be of same length.

Parameters

• bufferScaled (typing.Iterable[float]) – scaled parameters

• pscale (typing.Iterable[amici.amici.ParameterScaling]) – parameter scaling

• bufferUnscaled (typing.Iterable[float]) – unscaled parameters are written to the array

Return type

None