ott.solvers.linear.semidiscrete.SemidiscreteSolver.step#
- SemidiscreteSolver.step(rng, state, prob, *, compute_error=False, rng_error=None)[source]#
Perform one optimization step.
- Parameters:
rng (
Array) – Random seed used for sampling.state (
SemidiscreteState) – Semidiscrete state.prob (
SemidiscreteLinearProblem) – Semidiscrete linear problem.compute_error (
bool) – Whether to compute the marginal chi-squared error.rng_error (
Optional[Array]) – Random seed when computing the chi-squared error.
- Return type:
- Returns:
The updated state.