cottax.relational.consumers.execution.jax_solvers.optimistix module¶
Newton, Picard and BFGS: optimistix answering a RootFinder, a FixedPointIterator and a Minimiser.
- class Optimistix[D: DriverContract](rtol: float = 0.0001, atol: float = 0.0001, max_steps: int = 256)[source]¶
Bases:
DriverImplementation[D],Generic[D]An optimistix solve from a Start, reporting its Steps. The unknowns are optimistix’s y as they are, a tuple, and the context its args.
- needs: AbstractClassVar[frozenset[type[Role]]] = frozenset({<class 'cottax.core.driven.roles.Start'>})¶
- provides: AbstractClassVar[frozenset[type[Role]]] = frozenset({<class 'cottax.core.driven.roles.Steps'>})¶
- rtol: float = 0.0001¶
- atol: float = 0.0001¶
- max_steps: int = 256¶
- type Args = tuple[Posed, Context]¶
What each solver function is handed besides the unknowns: the static half, then the values.
- class Newton(rtol: float = 0.0001, atol: float = 0.0001, max_steps: int = 256)[source]¶
Bases:
Optimistix[RootFinder]optx.Newton on the equality residuals, as many numbers as the unknowns.
- solve[V: Place](posed: Posed[V], context: Context, data: ByRole)[source]¶
The optimistix solve answering posed: its function takes args=(posed, context).
- Return type:
Solution
- implements¶
alias of
RootFinder
- class Picard(rtol: float = 0.0001, atol: float = 0.0001, max_steps: int = 256)[source]¶
Bases:
Optimistix[FixedPointIterator]optx.FixedPointIteration: u <- g(u), no Jacobian taken.
- solve[V: Place](posed: Posed[V], context: Context, data: ByRole)[source]¶
The optimistix solve answering posed: its function takes args=(posed, context).
- Return type:
Solution
- implements¶
alias of
FixedPointIterator
- class BFGS(rtol: float = 0.0001, atol: float = 0.0001, max_steps: int = 256)[source]¶
Bases:
Optimistix[Minimiser]optx.BFGS on the one objective.