Function wodinFit

  • Begin a fit. This will evaluate the model pars.vary.length + 1 times, and then return a dfoptim.Simplex object. You can then call .run() to fit the model in one go (could block for a long time) or repeatedly call .step() until it returns true when it has converged.

    The data field of the returned value, both during a run via .result() and on convergence (perhaps via .run()) will have interpolated solutions available via an object containing:

    • names: the names of all traces returned by the model
    • pars: The full model parameters, as an object mapping strings to UserValue (i.e., suitable to pass back into an OdinModelConstructable object or wodinRun)
    • solutionAll: The solution of all series; an interpolating function as as would be returned by wodinRun
    • solutionFit: The solution to a just the modelled series being fit, as a single trace

    See FitResult

    Parameters

    • Model: OdinModelConstructable

      The model constructor

    • data: FitData

      The data to fit to; there will be one time and one data series within this

    • pars: FitPars

      Information about the parameters; which are to be varied, which to be fixed, and their starting values

    • modelledSeries: string

      Name of the series of data produced by the model that should be compared with the data.

    • controlODE: Partial<DopriControlParam>

      Control parameters to tune the integration

    • controlFit: Partial<SimplexControlParam>

      Control parameters to tune the optimisation

    Returns Simplex

Generated using TypeDoc