Bristlecone (bristlecone) | bristlecone

bristlecone

Bristlecone Module

Namespace: Bristlecone
Assembly: Bristlecone.dll

Main functionality of Bristlecone, including functions to scaffold `ModelSystem`s and for model-fitting (tests and real fits).

Nested modules

Modules	Description
Parallel	Wrappers for fitting functions that use `Array.Parallel` to run many analyses at once.

Functions and values

Function or value	Description
`Bristlecone.bootstrap engine endCondition bootstrapCount model series` Full Usage: `Bristlecone.bootstrap engine endCondition bootstrapCount model series` Parameters: engine : `EstimationEngine<'date, 'timespan, 'modelTimeUnit, MeasureOne>` - The estimation engine / fitting method endCondition : `EndCondition` - The end condition for each model fit bootstrapCount : `int` - Number of times to bootstrap data model : `ModelSystem<'modelTimeUnit>` - A model system / hypothesis to fit series : `Map<ShortCode, TimeSeries<float, 'date, 'timeunit, 'timespan>>` - Time-series to fit with model Returns: `EstimationResult<'date, 'timeunit, 'timespan> list` A list of estimation results (one for each bootstrap) for further analysis	Repeat a model fit many times, removing a single data point at random each time. engine : `EstimationEngine<'date, 'timespan, 'modelTimeUnit, MeasureOne>` The estimation engine / fitting method endCondition : `EndCondition` The end condition for each model fit bootstrapCount : `int` Number of times to bootstrap data model : `ModelSystem<'modelTimeUnit>` A model system / hypothesis to fit series : `Map<ShortCode, TimeSeries<float, 'date, 'timeunit, 'timespan>>` Time-series to fit with model Returns: `EstimationResult<'date, 'timeunit, 'timespan> list` A list of estimation results (one for each bootstrap) for further analysis
`Bristlecone.fit engine endCondition timeSeriesData model` Full Usage: `Bristlecone.fit engine endCondition timeSeriesData model` Parameters: engine : `EstimationEngine<'a, 'b, 'modelTimeUnit, 'u>` - An estimation engine configured and tested for the given model. endCondition : `EndCondition` - The condition at which optimisation should cease. timeSeriesData : `CodedMap<TimeSeries<float<'u>, 'a, 'c, 'b>>` - Time-series dataset that contains a series for each equation in the model system. model : `ModelSystem<'modelTimeUnit>` - A model system of equations, likelihood function, estimatible parameters, and optional measures. Returns: `EstimationResult<'a, 'c, 'b>` The result of the model-fitting procedure. If an error occurs, throws an exception.	Fit a time-series model to data. engine : `EstimationEngine<'a, 'b, 'modelTimeUnit, 'u>` An estimation engine configured and tested for the given model. endCondition : `EndCondition` The condition at which optimisation should cease. timeSeriesData : `CodedMap<TimeSeries<float<'u>, 'a, 'c, 'b>>` Time-series dataset that contains a series for each equation in the model system. model : `ModelSystem<'modelTimeUnit>` A model system of equations, likelihood function, estimatible parameters, and optional measures. Returns: `EstimationResult<'a, 'c, 'b>` The result of the model-fitting procedure. If an error occurs, throws an exception.
`Bristlecone.forDailyModel engine` Full Usage: `Bristlecone.forDailyModel engine` Parameters: engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<DateTime, TimeSpan, MeasureProduct<day, MeasureOne>, 'v>`	engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<DateTime, TimeSpan, MeasureProduct<day, MeasureOne>, 'v>`
`Bristlecone.forMonthlyModel engine` Full Usage: `Bristlecone.forMonthlyModel engine` Parameters: engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<DateTime, TimeSpan, MeasureProduct<month, MeasureOne>, 'v>`	engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<DateTime, TimeSpan, MeasureProduct<month, MeasureOne>, 'v>`
`Bristlecone.mkContinuous ()` Full Usage: `Bristlecone.mkContinuous ()` Parameters: () : `unit` Returns: `EstimationEngine<DateTime, TimeSpan, MeasureProduct<year, MeasureOne>, 'u>`	A basic estimation engine for ordinary differential equations, using a Nelder-Mead optimiser. () : `unit` Returns: `EstimationEngine<DateTime, TimeSpan, MeasureProduct<year, MeasureOne>, 'u>`
`Bristlecone.mkDiscrete ()` Full Usage: `Bristlecone.mkDiscrete ()` Parameters: () : `unit` Returns: `EstimationEngine<DateTime, TimeSpan, MeasureProduct<year, MeasureOne>, 'u>`	A basic estimation engine for discrete-time equations, using a Nelder-Mead optimiser. () : `unit` Returns: `EstimationEngine<DateTime, TimeSpan, MeasureProduct<year, MeasureOne>, 'u>`
`Bristlecone.oneStepAhead engine hypothesis preTransform timeSeries estimatedTheta` Full Usage: `Bristlecone.oneStepAhead engine hypothesis preTransform timeSeries estimatedTheta` Parameters: engine : `EstimationEngine<'date, 'a, 'modelTimeUnit, MeasureOne>` - The exact estimation engine used for the existing model fit hypothesis : `ModelSystem<'modelTimeUnit>` - The exact model system / hypothesis from which parameters have been already estimated preTransform : `Map<ShortCode, TimeSeries<float, 'date, 'timeunit, 'timespan>> -> CodedMap<TimeSeries<float<MeasureOne>, 'date, 'b, 'a>>` - A function that may transform each shorter time-series before prediction. This may be needed, for example, if there are custom start values that need to be configured in a complex way (e.g. for derived mesaurement variables). timeSeries : `Map<ShortCode, TimeSeries<float, 'date, 'timeunit, 'timespan>>` - The observed data to predict against. estimatedTheta : `ParameterPool` - A parameter pool containing already estimated parameters from model fitting step Returns: `Map<ShortCode, (TimeSeries<FitValue, 'date, 'timeunit, 'timespan> * NStepStatistics)>` A time-series for each variable containing a step-ahead prediction	Addresses the question: "How good am I at predicting the next data point?. Given fitted parameters, assesses how well the model predicts the next data point from each point in the time-series data. This approach can provide another indication of model performance. engine : `EstimationEngine<'date, 'a, 'modelTimeUnit, MeasureOne>` The exact estimation engine used for the existing model fit hypothesis : `ModelSystem<'modelTimeUnit>` The exact model system / hypothesis from which parameters have been already estimated preTransform : `Map<ShortCode, TimeSeries<float, 'date, 'timeunit, 'timespan>> -> CodedMap<TimeSeries<float<MeasureOne>, 'date, 'b, 'a>>` A function that may transform each shorter time-series before prediction. This may be needed, for example, if there are custom start values that need to be configured in a complex way (e.g. for derived mesaurement variables). timeSeries : `Map<ShortCode, TimeSeries<float, 'date, 'timeunit, 'timespan>>` The observed data to predict against. estimatedTheta : `ParameterPool` A parameter pool containing already estimated parameters from model fitting step Returns: `Map<ShortCode, (TimeSeries<FitValue, 'date, 'timeunit, 'timespan> * NStepStatistics)>` A time-series for each variable containing a step-ahead prediction
`Bristlecone.testModel engine settings model` Full Usage: `Bristlecone.testModel engine settings model` Parameters: engine : `EstimationEngine<'date, 'timespan, 'modelTimeUnit, 'state>` - An estimation engine containing the method used for model-fitting. settings : `TestSettings<'state, 'date, 'yearUnit, 'timespan>` - Test settings that define how the test will be conducted. model : `ModelSystem<'modelTimeUnit>` - The model system to test against the estimation engine. Returns: `TestResult<'date, 'yearUnit, 'timespan, MeasureProduct<state, MeasureOne>>` A test result that indicates differences between the expected and actual fit.	Test that the specified estimation engine can correctly estimate known parameters. Random parameter sets are generated from the given model system. engine : `EstimationEngine<'date, 'timespan, 'modelTimeUnit, 'state>` An estimation engine containing the method used for model-fitting. settings : `TestSettings<'state, 'date, 'yearUnit, 'timespan>` Test settings that define how the test will be conducted. model : `ModelSystem<'modelTimeUnit>` The model system to test against the estimation engine. Returns: `TestResult<'date, 'yearUnit, 'timespan, MeasureProduct<state, MeasureOne>>` A test result that indicates differences between the expected and actual fit.
`Bristlecone.tryFit engine endCondition observedSeries model` Full Usage: `Bristlecone.tryFit engine endCondition observedSeries model` Parameters: engine : `EstimationEngine<'date, 'timespan, 'modelTimeUnit, 'stateUnit>` - The engine encapsulates all settings that form part of the estimation method. Importantly, this includes the random number generator used for all stages of the analysis; if this is set using a fixed seed, the result will be reproducable. endCondition : `EndCondition` - You must specify a stopping condition, after which the optimisation process will cease. Bristlecone includes built-in end conditions in the `Bristlecone.Optimisation.EndConditions` module. observedSeries : `CodedMap<TimeSeries<float<'stateUnit>, 'date, 'yearType, 'timespan>>` model : `ModelSystem<'modelTimeUnit>` - Returns: `Result<EstimationResult<'date, 'yearType, 'timespan>, string>`	Fit a time-series model to data. Please note: it is strongly recommended that you test that the given `EstimationEngine` can correctly identify known parameters for your model. Refer to the `Bristlecone.testModel` function, which can be used to generate known data and complete this process. engine : `EstimationEngine<'date, 'timespan, 'modelTimeUnit, 'stateUnit>` The engine encapsulates all settings that form part of the estimation method. Importantly, this includes the random number generator used for all stages of the analysis; if this is set using a fixed seed, the result will be reproducable. endCondition : `EndCondition` You must specify a stopping condition, after which the optimisation process will cease. Bristlecone includes built-in end conditions in the `Bristlecone.Optimisation.EndConditions` module. observedSeries : `CodedMap<TimeSeries<float<'stateUnit>, 'date, 'yearType, 'timespan>>` model : `ModelSystem<'modelTimeUnit>` Returns: `Result<EstimationResult<'date, 'yearType, 'timespan>, string>`
`Bristlecone.tryTestModel engine settings model` Full Usage: `Bristlecone.tryTestModel engine settings model` Parameters: engine : `EstimationEngine<'date, 'timespan, 'modelTimeUnit, 'state>` - settings : `TestSettings<'state, 'date, 'timeunit, 'timespan>` - model : `ModelSystem<'modelTimeUnit>` - Returns: `Result<TestResult<'date, 'timeunit, 'timespan, MeasureProduct<state, MeasureOne>>, string>` A test result that indicates the error structure. It is wrapped in an F# Result, indicating if the procedure was successful or not.	Tests that the specified estimation engine can correctly estimate known parameters given specfici test settings. Random parameter sets and resultant fake time-series data are generated for the model system by using the rules and noise generation settings in the stated test settings. engine : `EstimationEngine<'date, 'timespan, 'modelTimeUnit, 'state>` settings : `TestSettings<'state, 'date, 'timeunit, 'timespan>` model : `ModelSystem<'modelTimeUnit>` Returns: `Result<TestResult<'date, 'timeunit, 'timespan, MeasureProduct<state, MeasureOne>>, string>` A test result that indicates the error structure. It is wrapped in an F# Result, indicating if the procedure was successful or not.
`Bristlecone.withConditioning c engine` Full Usage: `Bristlecone.withConditioning c engine` Parameters: c : `Conditioning<'u>` engine : `EstimationEngine<'a, 'b, 'v, 'u>` Returns: `EstimationEngine<'a, 'b, 'v, 'u>`	Choose how the start point is chosen when solving the model system c : `Conditioning<'u>` engine : `EstimationEngine<'a, 'b, 'v, 'u>` Returns: `EstimationEngine<'a, 'b, 'v, 'u>`
`Bristlecone.withContinuousTime t engine` Full Usage: `Bristlecone.withContinuousTime t engine` Parameters: t : `IntegrationRoutine` engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`	Use a custom integration method t : `IntegrationRoutine` engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`
`Bristlecone.withCustomOptimisation optim engine` Full Usage: `Bristlecone.withCustomOptimisation optim engine` Parameters: optim : `Optimiser` engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`	optim : `Optimiser` engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`
`Bristlecone.withGradientDescent engine` Full Usage: `Bristlecone.withGradientDescent engine` Parameters: engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`	engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`
`Bristlecone.withOutput out engine` Full Usage: `Bristlecone.withOutput out engine` Parameters: out : `WriteOut` - engine : `EstimationEngine<'a, 'b, 'u, 'v>` - Returns: `EstimationEngine<'a, 'b, 'u, 'v>`	Substitute a specific logger into out : `WriteOut` engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`
`Bristlecone.withSeed seed engine` Full Usage: `Bristlecone.withSeed seed engine` Parameters: seed : `int` engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`	Use a mersenne twister random number generator with a specific seed. seed : `int` engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`
`Bristlecone.withTimeConversion fn engine` Full Usage: `Bristlecone.withTimeConversion fn engine` Parameters: fn : `ResolutionToModelUnits<'d2, 'timespan2, 'modelTimeUnit>` engine : `EstimationEngine<'d, 'o1, 'o2, 'u>` Returns: `EstimationEngine<'d2, 'timespan2, 'modelTimeUnit, 'u>`	fn : `ResolutionToModelUnits<'d2, 'timespan2, 'modelTimeUnit>` engine : `EstimationEngine<'d, 'o1, 'o2, 'u>` Returns: `EstimationEngine<'d2, 'timespan2, 'modelTimeUnit, 'u>`
`Bristlecone.withTunedMCMC tuning engine` Full Usage: `Bristlecone.withTunedMCMC tuning engine` Parameters: tuning : `TuneStep seq` engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`	tuning : `TuneStep seq` engine : `EstimationEngine<'a, 'b, 'u, 'v>` Returns: `EstimationEngine<'a, 'b, 'u, 'v>`