Package jline.solvers.tr

Class LcStrategy

java.lang.Object

jline.solvers.tr.LcStrategy

All Implemented Interfaces:

TransformSolve.Strategy

public class LcStrategy
extends Object
implements TransformSolve.Strategy

LOAD CONCEALMENT as a model transformation, and the first ITERATED strategy of TransformSolve.

Birman and Kogan (Stochastic Models 8(3):543-563, 1992), Algorithm 2. The saddle point analysis of their Corollary 2 shows that chain l may be solved on its own provided every station is slowed by the residual capacity the other chains leave it, A_i = 1 - sum_{k!=l} L(i,k) X_k, so that chain l sees the concealed demand L(i,l)/A_i.

WHAT THIS ADDS OVER THE KERNEL. Pfqn_bk.pfqn_bklc solves each single-chain subproblem on a DEMAND VECTOR with the inner solve hard-wired to MVA or the uniform expansion. Here the subproblem is a real single-class Network, so the inner solve is the CALLER'S OWN solver, which is what makes the concealment approximation measurable rather than merely asserted.

IT IS NOT A STRICTLY BETTER LC. The kernel sees only L; the chain aggregation refits the chain service law to two moments. On a product-form model they coincide; off it they are different approximations.

THE TOLERANCE IS FIXED AT 1e-10 and is deliberately not options.iter_tol: a looser one stops the sweep at a different iteration in each codebase.

Mirrors MATLAB solver_tr_lc_analyzer.m and python transform_driver._lc_strategy.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static final class	LcStrategy.LcContext	Carries the concealment state across the sweep.

Field Summary

Fields

Modifier and Type	Field	Description
static final double	TOL	The fixed convergence tolerance; a parity requirement, not a knob.

Constructor Summary

Constructors

Constructor	Description
LcStrategy()

Method Summary

Modifier and Type	Method	Description
boolean	converged(TransformSolve.Expanded ctx, List<TransformSolve.Inner> res, int it)	True when the sweep may stop.
TransformSolve.Expanded	couple(TransformSolve.Expanded ctx, List<TransformSolve.Inner> res, int e)	Updates the subproblems from the sweep so far.
TransformSolve.Expanded	expand(Network model, NetworkStruct sn, SolverOptions options)	Builds the subproblems and the context the lift will need.
TransformSolve.Lifted	lift(TransformSolve.Expanded ctx, List<TransformSolve.Inner> res)	Maps the subproblem metrics back onto the original model.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

TOL

public static final double TOL

The fixed convergence tolerance; a parity requirement, not a knob.

See Also:

• Constant Field Values

Constructor Details

LcStrategy

public LcStrategy()

Method Details

expand

public TransformSolve.Expanded expand(Network model,
 NetworkStruct sn,
 SolverOptions options)

Description copied from interface: TransformSolve.Strategy

Builds the subproblems and the context the lift will need.

Specified by:

expand in interface TransformSolve.Strategy

couple

public TransformSolve.Expanded couple(TransformSolve.Expanded ctx,
 List<TransformSolve.Inner> res,
 int e)

Description copied from interface: TransformSolve.Strategy

Updates the subproblems from the sweep so far. Called only by an iterated strategy, immediately after each subproblem's solve, which is what makes the coupling Gauss-Seidel.

Specified by:

couple in interface TransformSolve.Strategy

converged

public boolean converged(TransformSolve.Expanded ctx,
 List<TransformSolve.Inner> res,
 int it)

Description copied from interface: TransformSolve.Strategy

True when the sweep may stop. Called only by an iterated strategy.

Specified by:

converged in interface TransformSolve.Strategy

lift

public TransformSolve.Lifted lift(TransformSolve.Expanded ctx,
 List<TransformSolve.Inner> res)

Description copied from interface: TransformSolve.Strategy

Maps the subproblem metrics back onto the original model.

Specified by:

lift in interface TransformSolve.Strategy