Objective (bristlecone) | bristlecone

bristlecone

Objective Module

Namespace: Bristlecone
Assembly: Bristlecone.dll

Configures a single function that represents a model and its likelihood when fit to time-series data.

Functions and values

Function or value	Description
`Objective.accessorFromRealVector compiled thetaReal` Full Usage: `Objective.accessorFromRealVector compiled thetaReal` Parameters: compiled : `CompiledTransforms<'space>` thetaReal : `TypedTensor<Vector, MeasureProduct<parameter, MeasureOne>>` Returns: `ParameterValueAccessor`	compiled : `CompiledTransforms<'space>` thetaReal : `TypedTensor<Vector, MeasureProduct<parameter, MeasureOne>>` Returns: `ParameterValueAccessor`
`Objective.compiledFromConfig config` Full Usage: `Objective.compiledFromConfig config` Parameters: config : `AnyOptimiserConfig` Returns: `CompiledTransforms<MeasureProduct<optim-space, MeasureOne>>`	config : `AnyOptimiserConfig` Returns: `CompiledTransforms<MeasureProduct<optim-space, MeasureOne>>`
`Objective.create negLogLikFn measures solver config observed` Full Usage: `Objective.create negLogLikFn measures solver config observed` Parameters: negLogLikFn : `Likelihood<MeasureProduct<state, MeasureOne>>` measures : `CodedMap<Measurement<MeasureProduct<state, MeasureOne>>>` solver : `ConfiguredSolver` config : `AnyOptimiserConfig` observed : `CodedMap<float<MeasureProduct<state, MeasureOne>>[]>` Returns: `Objective`	Computes measurement variables and appends to expected data. Requires: - A continuous-time solver (for differential-based equations) - A discrete-time solver (for if measurement / computed variables are present) - Observed data (for calculating likelihood) negLogLikFn : `Likelihood<MeasureProduct<state, MeasureOne>>` measures : `CodedMap<Measurement<MeasureProduct<state, MeasureOne>>>` solver : `ConfiguredSolver` config : `AnyOptimiserConfig` observed : `CodedMap<float<MeasureProduct<state, MeasureOne>>[]>` Returns: `Objective`
`Objective.createPredictor measures solver config` Full Usage: `Objective.createPredictor measures solver config` Parameters: measures : `CodedMap<Measurement<MeasureProduct<state, MeasureOne>>>` solver : `ConfiguredSolver` config : `AnyOptimiserConfig` Returns: `TypedTensor<Vector, MeasureProduct<optim-space, MeasureOne>> -> Map<ShortCode, TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>`	measures : `CodedMap<Measurement<MeasureProduct<state, MeasureOne>>>` solver : `ConfiguredSolver` config : `AnyOptimiserConfig` Returns: `TypedTensor<Vector, MeasureProduct<optim-space, MeasureOne>> -> Map<ShortCode, TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>`
`Objective.measure measures parameters expectedDynamic initialConditions` Full Usage: `Objective.measure measures parameters expectedDynamic initialConditions` Parameters: measures : `CodedMap<Measurement<MeasureProduct<state, MeasureOne>>>` parameters : `TypedTensor<Vector, MeasureProduct<parameter, MeasureOne>>` expectedDynamic : `CodedMap<TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>` initialConditions : `CodedMap<TypedTensor<Scalar, MeasureProduct<state, MeasureOne>>>` Returns: `CodedMap<TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>`	Compute the system's `Measures` from the dynamic variables produced by the solver. All operations happen in Tensor-space. Initial conditions (t0) are added to the front of the predictions to enable previous value lookup where needed. measures : `CodedMap<Measurement<MeasureProduct<state, MeasureOne>>>` parameters : `TypedTensor<Vector, MeasureProduct<parameter, MeasureOne>>` expectedDynamic : `CodedMap<TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>` initialConditions : `CodedMap<TypedTensor<Scalar, MeasureProduct<state, MeasureOne>>>` Returns: `CodedMap<TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>`
`Objective.pairObservationsToExpected observed expected` Full Usage: `Objective.pairObservationsToExpected observed expected` Parameters: observed : `Map<ShortCode, TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>` expected : `Map<ShortCode, TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>` Returns: `CodedMap<SeriesPair<MeasureProduct<state, MeasureOne>>>`	Pairs observed time series to predicted series for dynamic variables only. Environmental forcings and hidden variables are removed. observed : `Map<ShortCode, TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>` expected : `Map<ShortCode, TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>` Returns: `CodedMap<SeriesPair<MeasureProduct<state, MeasureOne>>>`
`Objective.predict solver measures parameters` Full Usage: `Objective.predict solver measures parameters` Parameters: solver : `TypedTensor<Vector, MeasureProduct<parameter, MeasureOne>> -> CodedMap<TypedTensor<Vector, MeasureProduct<state, MeasureOne>>> * CodedMap<TypedTensor<Scalar, MeasureProduct<state, MeasureOne>>>` measures : `CodedMap<Measurement<MeasureProduct<state, MeasureOne>>>` parameters : `TypedTensor<Vector, MeasureProduct<parameter, MeasureOne>>` Returns: `Map<ShortCode, TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>`	solver : `TypedTensor<Vector, MeasureProduct<parameter, MeasureOne>> -> CodedMap<TypedTensor<Vector, MeasureProduct<state, MeasureOne>>> * CodedMap<TypedTensor<Scalar, MeasureProduct<state, MeasureOne>>>` measures : `CodedMap<Measurement<MeasureProduct<state, MeasureOne>>>` parameters : `TypedTensor<Vector, MeasureProduct<parameter, MeasureOne>>` Returns: `Map<ShortCode, TypedTensor<Vector, MeasureProduct<state, MeasureOne>>>`
`Objective.prependInitialConditions initial expected` Full Usage: `Objective.prependInitialConditions initial expected` Parameters: initial : `Map<'a, TypedTensor<Scalar, 'u>>` expected : `Map<'a, TypedTensor<Vector, 'u>>` Returns: `Map<'a, TypedTensor<Vector, 'u>>`	initial : `Map<'a, TypedTensor<Scalar, 'u>>` expected : `Map<'a, TypedTensor<Vector, 'u>>` Returns: `Map<'a, TypedTensor<Vector, 'u>>`