Package jline.api.fj

Class FJ_dispersion

java.lang.Object

jline.api.fj.FJ_dispersion

public final class FJ_dispersion
extends Object

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static final class	FJ_dispersion.FJDelayOptResult	[d, Edisp, Emax] of fj_delay_opt.
static final class	FJ_dispersion.FJDispersionResult	[Edisp, Emax, Emin] of fj_dispersion.

Method Summary

Modifier and Type	Method	Description
static FJ_dispersion.FJDelayOptResult	fj_delay_opt(int[] shape, double[] rate)
static FJ_dispersion.FJDelayOptResult	fj_delay_opt(int[] shape, double[] rate, int maxsweeps, double dtol, int npanels)	The deterministic delays that minimise the mean dispersion.
static FJ_dispersion.FJDispersionResult	fj_dispersion(int[] shape, double[] rate)
static FJ_dispersion.FJDispersionResult	fj_dispersion(int[] shape, double[] rate, double[] d, double tol, int npanels)	Mean subtask dispersion of a split-merge system whose branches are shifted Erlangs, the equivalent used in the delay-scheduling construction: E[D_d] = integral_0^inf [ 1 - prod_i F_i(x-d_i) - prod_i (1-F_i(x-d_i)) ] dx.

Methods inherited from class java.lang.Object

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

Method Details

fj_dispersion

public static FJ_dispersion.FJDispersionResult fj_dispersion(int[] shape,
 double[] rate,
 double[] d,
 double tol,
 int npanels)

Mean subtask dispersion of a split-merge system whose branches are shifted Erlangs, the equivalent used in the delay-scheduling construction: E[D_d] = integral_0^inf [ 1 - prod_i F_i(x-d_i) - prod_i (1-F_i(x-d_i)) ] dx. That integrand is non-negative and vanishes at both ends; the difference of the two products printed in the survey is not the dispersion and can go negative.

fj_dispersion

public static FJ_dispersion.FJDispersionResult fj_dispersion(int[] shape,
 double[] rate)

fj_delay_opt

public static FJ_dispersion.FJDelayOptResult fj_delay_opt(int[] shape,
 double[] rate,
 int maxsweeps,
 double dtol,
 int npanels)

The deterministic delays that minimise the mean dispersion. Holding back a fast branch costs little at the last completion and buys a great deal at the first, so the minimiser is generally interior and strictly positive on every branch but the slowest. Cyclic coordinate descent with a golden section line search on each coordinate: deterministic, derivative-free, and the same sequence of evaluations in all four codebases. Adding a constant to every delay shifts both order statistics equally, so the representative with min(d) = 0 is returned.

fj_delay_opt

public static FJ_dispersion.FJDelayOptResult fj_delay_opt(int[] shape,
 double[] rate)