ott.solvers.linear.semidiscrete.SemidiscreteOutput

class ott.solvers.linear.semidiscrete.SemidiscreteOutput(g, prob, it=None, losses=None, errors=None, converged=None)

Output of the SemidiscreteSolver.

Parameters:

• g (Array) – Dual potential.

• prob (SemidiscreteLinearProblem) – Semidiscrete OT problem.

• it (Optional[int]) – Final iteration number.

• losses (Optional[Array]) – Dual losses.

• errors (Optional[Array]) – Marginal deviation errors.

• converged (Optional[bool]) – Whether the solver converged.

Methods

marginal_chi2_error(rng, *, num_iters, ...)	Compute the marginal chi-squared error.
sample(rng, num_samples, *[, epsilon])	Sample a point cloud and compute the OT solution.
to_dataloader(rng, batch_size, **kwargs)	Create a semidiscrete dataloader.
to_dual_potentials([epsilon])	Compute the dual potential function \(f\).

Attributes

converged
errors
geom	Semidiscrete geometry.
it
losses
g
prob