-
Notifications
You must be signed in to change notification settings - Fork 0
/
input.in
148 lines (143 loc) · 6.45 KB
/
input.in
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
/*Input file*/
RUN = 30
DIM = 250
FUNCTION = 0
KMAX = 5
TMAX = 500000
Q = 2.8
RADII = 0.1
RADII_T_FLAG = 1
AVAL_MAX = 2000
RHO = 0.5
EPSILON = 1E-4
METHOD = 6
P=0
VNS_POP = 50
ECO_STEP = 5
EVO_STEP = 50
G_MAX = 10000
PC = 0.9
/* RUN is the number of times the ECO alg will run.*/
/* DIM is the number of problem variables. */
/* FUNCTION represents the problem to be solved. The current options are:*/
/*N Name Range Min */
/*0: Rastrigin [ -5.12, 5.12] 0 */
/*1: Schaffer F7 [-100.00, 100.00] 0 */
/*2: Griewank [-600.00, 600.00] 0 */
/*3: Ackley [ -32.00, 32.00] 0 */
/*4: Rosenbrock [ -30.00, 30.00] 0 */
/*5: Sphere [-100.00, 100.00] 0 */
/*6: Molecular Potential Energy [ 0.00, 5.00] 0 */
/*7: Schaffer F6 [-100.00, 100.00] 0 */
/*8: Generalized Schwefels 2.26 [-500.00, 500.00] 0 */
/*9: Step [ -100, 100] 0 */
/*11: Generalized Penalized function #1 [ -50, 50] 1.57e-032 */
/*12: Levy [ -10.00, 10.00] 0 */
/*13: Zakharov [ -5.00, 10.00] 0 */
/*14: Egg Holder [-512.00, 512.00] ? */
/*15: Holzman [ -10.00, 10.00] 0 */
/*16: Michalewitz [ 0.00, PI] -0.966*n */
/*17: Generalized Penalized function #2 [ -50, 50] 1.34e-032 */
/*18: Powell [ -10.00, 10.00] 0 */
/*19: Rana [-512.00, 512.00] ? */
/*20: Shubert [ -4.00, 5.00] -24.06 */
/*21: StretchedV [-10.00, 10.00] 0 */
/*22: Multimod [-10.00, 10.00] 0 */
/*23: Schwefel's function 2.22 [ -10, 10] 0 */
/*25: Shifited Sphere [ -100, 100] -450 */
/*26: Shifted Schwefel Problem 2.21 [ -100, 100] -450 */
/*27: Shifted Rosenbrock [ -100, 100] 390 */
/*28: Shifted Rastrigin [ -5.12, 5.12] -330 */
/*30: Shifted Ackley [ -32.00, 32.00] -180 */
/*31: Shifted Schwefel 2.22 [ -10.00, 10.00] -140 */
/*32: Shifted Schwefel 1.2 [-64.536, 65.536] 0 */
/*33: Shifted Extended_f10 [ -100, 100] 0 */
/*34: Shifted Bohachevsky [ -15, 15] 0 */
/*35: Shifted Schaffer [ -100, 100] 0 */
/*36: Hybrid 1 [ -100, 100] 0 */
/*37: Hybrid 2 [ -5, 5] 0 */
/*38: Hybrid 3 [ -10, 10] 0 */
/*39: Hybrid 4 [ -100, 100] 0 */
/*40: Hybrid 5 [ -100, 100] 0 */
/*41: Hybrid 6 [ -100, 100] 0 */
/*42: Hybrid 7 [ -5, 5] 0 */
/*43: Hybrid 8 [ -10, 10] 0 */
/****VNS PARAMETERS****/
/*K is the number of neighbourhood*/
/*TMAX is the number os iterations*/
/*Q is the geometric progresion rate*/
/*RADII is the radio aplied, the initial value of geometric progresion Q*/
/*RADII_T_FLAG is the flag for radii type test:
0 => r[k-1] < pk < r[k]*/
1 => pk < r[k]*/
/*AVAL_MAX for BVNS, the number of avaliation used in local search
/*METHOD the algorithm to run*/
/* 2 => HJ*/
/* 3 => BVNS*/
/* 4 => RVNS*/
/* 5 => BVNS_NM */
/* 6 => PRVNS*/
/* 7 => NM*/
/*P the metric used:*/
/*0 => l_inf (shebchev)*/
/*1 => l1 (manhathan)*/
/*2 => l2 (euclidian)*/
/* VNS_POP population size for PVNS. */
/* G_MAX the max generation for PVNS, must be balanced with AVAL_MAX. G_MAX = AVAL_MAX/VNS_POP.
/* PC Cross-Over probability used in PVNS */
/****HOOKE AND JEEVES PARAMETERS****/
/*ITERMAX = */
/*RHO = */
/*EPSILON = */
/* rho {a double} This is a user-supplied convergence */
/* parameter (more detail below), which should be */
/* set to a value between 0.0 and 1.0. Larger */
/* values of rho give greater probability of */
/* convergence on highly nonlinear functions, at a */
/* cost of more function evaluations. Smaller */
/* values of rho reduces the number of evaluations */
/* (and the program running time), but increases */
/* the risk of nonconvergence. See below. */
/* epsilon {a double} This is the criterion for halting */
/* the search for a minimum. When the algorithm */
/* begins to make less and less progress on each */
/* iteration, it checks the halting criterion: if */
/* the stepsize is below epsilon, terminate the */
/* iteration and return the current best estimate */
/* of the minimum. Larger values of epsilon (such */
/* as 1.0e-4) give quicker running time, but a */
/* less accurate estimate of the minimum. Smaller */
/* values of epsilon (such as 1.0e-7) give longer */
/* running time, but a more accurate estimate of */
/* the minimum. */
/* itermax {an integer} A second, rarely used, halting */
/* criterion. If the algorithm uses >= itermax */
/* iterations, halt. */
/* rho, the algorithm convergence control */
/* The algorithm works by taking "steps" from one estimate of */
/* a minimum, to another (hopefully better) estimate. Taking */
/* big steps gets to the minimum more quickly, at the risk of */
/* "stepping right over" an excellent point. The stepsize is */
/* controlled by a user supplied parameter called rho. At each */
/* iteration, the stepsize is multiplied by rho (0 < rho < 1), */
/* so the stepsize is successively reduced. */
/* Small values of rho correspond to big stepsize changes, */
/* which make the algorithm run more quickly. However, there */
/* is a chance (especially with highly nonlinear functions) */
/* that these big changes will accidentally overlook a */
/* promising search vector, leading to nonconvergence. */
/* Large values of rho correspond to small stepsize changes, */
/* which force the algorithm to carefully examine nearby points */
/* instead of optimistically forging ahead. This improves the */
/* probability of convergence. */
/* The stepsize is reduced until it is equal to (or smaller */
/* than) epsilon. So the number of iterations performed by */
/* Hooke-Jeeves is determined by rho and epsilon: */
/* rho**(number_of_iterations) = epsilon */
/* In general it is a good idea to set rho to an aggressively */
/* small value like 0.5 (hoping for fast convergence). Then, */
/* if the user suspects that the reported minimum is incorrect */
/* (or perhaps not accurate enough), the program can be run */
/* again with a larger value of rho such as 0.85, using the */
/* result of the first minimization as the starting guess to */
/* begin the second minimization. */