Low‑level compiled likelihood Works directly with a parameter tensor (real space).

The domain is fine to be float-based, as it is only used to initialise the optimisation routine. Represents the bounds and any constraint.

Determines if the end has been reached based on a list of tupled Solutions with their iteration number.

Model equations for estimation may be require time to be in indexed form (i.e. common across models and data). Parameter values are required in 'real' parameter units rather than (transformed) optimisation space.

An objective function that can be optimised within an optimisation routine.

Reasons optimisation may stop.

A parameterised RHS — parameters already bound. This is what the integration routine actually steps.

A point in optimisation-space. Optim-space is tensor-based, so all points are tensor vectors representing the parameters.

A function that, given parameters, produces a parameterised RHS for the ODE system. This is the output of the static solver setup.

Represents an external logging function.