Skip to content

Allowing fixed parameters (parameters that don't vary during optimization) #107

@atrophiedbrain

Description

@atrophiedbrain

When performing parameter estimation of a system of ODEs, one may need to pass in data that is necessary as part of the ODEFunction calculation but is not actually a parameter that needs to be optimized.

At the moment, one can declare these values as const and reference them in the ODEFunction or pack them into a tuple that is passed to the ODEFunction.

I note here in DifferentialEquations.jl, boundary information is passed at the optimizer level.

In an attempt to allow the p tuple passed to the ODEFunction to be divided up, I propose the following:

Could another tuple be passed at the optimizer level with boolean values to denote whether the corresponding entry in the packed p tuple should vary?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions