ott.solvers.linear.sinkhorn_lr.LRSinkhornOutput

class ott.solvers.linear.sinkhorn_lr.LRSinkhornOutput(q, r, g, costs, errors, ot_prob, epsilon, inner_iterations, reg_ot_cost=None)

Transport interface for a low-rank Sinkhorn solution.

Methods

apply(inputs[, axis])	Apply the transport to a array; axis=1 for its transpose.
compute_reg_ot_cost(ot_prob[, use_danskin])	param ot_prob:
cost_at_geom(other_geom)	Return OT cost for current solution, evaluated at any cost matrix.
count(value, /)	Return number of occurrences of value.
index(value[, start, stop])	Return first index of value.
marginal(axis)	param axis:
set(**kwargs)	Return a copy of self, with potential overwrites.
set_cost(ot_prob, lse_mode[, use_danskin])	param ot_prob:
transport_cost_at_geom(other_geom)	Return (by recomputing it) bare transport cost of current solution.

Attributes

a
b
converged
costs	Alias for field number 3
epsilon	Alias for field number 6
errors	Alias for field number 4
g	Alias for field number 2
geom
inner_iterations	Alias for field number 7
matrix	Transport matrix if it can be instantiated.
n_iters
ot_prob	Alias for field number 5
primal_cost	Return (by recomputing it) transport cost of current solution.
q	Alias for field number 0
r	Alias for field number 1
reg_ot_cost	Alias for field number 8
transport_mass	Sum of transport matrix.

Parameters:

• q (Array) –

• r (Array) –

• g (Array) –

• costs (Array) –

• errors (Array) –

• ot_prob (LinearProblem) –

• epsilon (float) –

• inner_iterations (int) –

• reg_ot_cost (float | None) –