ViewVC Help
View File | Revision Log | Show Annotations | Download File | Root Listing
root/jsr166/jsr166/src/test/loops/TSPExchangerTest.java
Revision: 1.7
Committed: Thu Oct 29 23:11:03 2009 UTC (14 years, 6 months ago) by jsr166
Branch: MAIN
Changes since 1.6: +1 -1 lines
Log Message:
typos

File Contents

# User Rev Content
1 dl 1.1 /*
2 dl 1.2 * Written by Doug Lea and Bill Scherer with assistance from members
3     * of JCP JSR-166 Expert Group and released to the public domain, as
4     * explained at http://creativecommons.org/licenses/publicdomain
5 dl 1.1 */
6    
7 dl 1.2 import java.util.*;
8 dl 1.1 import java.util.concurrent.*;
9     import java.util.concurrent.atomic.*;
10     import java.util.concurrent.locks.*;
11    
12 dl 1.2 /**
13     * A parallel Traveling Salesperson Problem (TSP) program based on a
14 dl 1.4 * genetic algorithm using an Exchanger. A population of chromosomes is
15     * distributed among "subpops". Each chromosomes represents a tour,
16     * and its fitness is the total tour length.
17 jsr166 1.6 *
18 dl 1.4 * A set of worker threads perform updates on subpops. The basic
19     * update step is:
20 dl 1.2 * <ol>
21 dl 1.4 * <li> Select a breeder b from the subpop
22 dl 1.2 * <li> Create a strand of its tour with a random starting point and length
23 jsr166 1.6 * <li> Offer the strand to the exchanger, receiving a strand from
24 dl 1.4 * another subpop
25 jsr166 1.6 * <li> Combine b and the received strand using crossing function to
26 dl 1.2 * create new chromosome c.
27 dl 1.4 * <li> Replace a chromosome in the subpop with c.
28 dl 1.2 * </ol>
29     *
30 dl 1.4 * This continues for a given number of generations per subpop.
31     * Because there are normally more subpops than threads, each worker
32     * thread performs small (randomly sized) run of updates for one
33     * subpop and then selects another. A run continues until there is at
34     * most one remaining thread performing updates.
35     *
36 dl 1.2 * See below for more details.
37     */
38 dl 1.1 public class TSPExchangerTest {
39 dl 1.3 static final int NCPUS = Runtime.getRuntime().availableProcessors();
40    
41 dl 1.4 /** Runs start with two threads, increasing by two through max */
42     static final int DEFAULT_MAX_THREADS = Math.max(4, NCPUS + NCPUS/2);
43    
44     /** The number of replication runs per thread value */
45     static final int DEFAULT_REPLICATIONS = 3;
46    
47     /** If true, print statistics in SNAPSHOT_RATE intervals */
48     static boolean verbose = true;
49     static final long SNAPSHOT_RATE = 10000; // in milliseconds
50 dl 1.2
51     /**
52     * The problem size. Each city is a random point. The goal is to
53     * find a tour among them with smallest total Euclidean distance.
54     */
55     static final int DEFAULT_CITIES = 144;
56    
57 jsr166 1.6 // Tuning parameters.
58 dl 1.2
59     /**
60 dl 1.4 * The number of chromosomes per subpop. Must be a power of two.
61 dl 1.2 *
62     * Smaller values lead to faster iterations but poorer quality
63     * results
64     */
65 jsr166 1.6 static final int DEFAULT_SUBPOP_SIZE = 32;
66 dl 1.1
67 dl 1.2 /**
68 dl 1.4 * The number of iterations per subpop. Convergence appears
69 dl 1.2 * to be roughly proportional to #cities-squared
70     */
71     static final int DEFAULT_GENERATIONS = DEFAULT_CITIES * DEFAULT_CITIES;
72    
73     /**
74 dl 1.4 * The number of subpops. The total population is #subpops * subpopSize,
75 dl 1.2 * which should be roughly on the order of #cities-squared
76     *
77     * Smaller values lead to faster total runs but poorer quality
78     * results
79     */
80 dl 1.4 static final int DEFAULT_NSUBPOPS = DEFAULT_GENERATIONS / DEFAULT_SUBPOP_SIZE;
81 dl 1.2
82     /**
83     * The minimum length for a random chromosome strand.
84     * Must be at least 1.
85     */
86     static final int MIN_STRAND_LENGTH = 3;
87    
88     /**
89 jsr166 1.5 * The probability mask value for creating random strands,
90 dl 1.2 * that have lengths at least MIN_STRAND_LENGTH, and grow
91 jsr166 1.7 * with exponential decay 2^(-(1/(RANDOM_STRAND_MASK + 1)
92 dl 1.2 * Must be 1 less than a power of two.
93     */
94     static final int RANDOM_STRAND_MASK = 7;
95    
96     /**
97 jsr166 1.5 * Probability control for selecting breeders.
98 dl 1.2 * Breeders are selected starting at the best-fitness chromosome,
99 jsr166 1.5 * with exponentially decaying probability
100 jsr166 1.6 * 1 / (subpopSize >>> BREEDER_DECAY).
101 dl 1.2 *
102     * Larger values usually cause faster convergence but poorer
103     * quality results
104     */
105     static final int BREEDER_DECAY = 1;
106    
107     /**
108 jsr166 1.5 * Probability control for selecting dyers.
109 dl 1.2 * Dyers are selected starting at the worst-fitness chromosome,
110 jsr166 1.5 * with exponentially decaying probability
111 dl 1.4 * 1 / (subpopSize >>> DYER_DECAY)
112 dl 1.2 *
113     * Larger values usually cause faster convergence but poorer
114     * quality results
115     */
116     static final int DYER_DECAY = 1;
117    
118     /**
119     * The set of cities. Created once per program run, to
120     * make it easier to compare solutions across different runs.
121     */
122 jsr166 1.6 static CitySet cities;
123 dl 1.1
124     public static void main(String[] args) throws Exception {
125 dl 1.2 int maxThreads = DEFAULT_MAX_THREADS;
126 dl 1.1 int nCities = DEFAULT_CITIES;
127 dl 1.4 int subpopSize = DEFAULT_SUBPOP_SIZE;
128 dl 1.2 int nGen = nCities * nCities;
129 dl 1.4 int nSubpops = nCities * nCities / subpopSize;
130     int nReps = DEFAULT_REPLICATIONS;
131 dl 1.1
132     try {
133 dl 1.2 int argc = 0;
134 dl 1.1 while (argc < args.length) {
135     String option = args[argc++];
136 dl 1.2 if (option.equals("-c")) {
137 dl 1.1 nCities = Integer.parseInt(args[argc]);
138 dl 1.2 nGen = nCities * nCities;
139 dl 1.4 nSubpops = nCities * nCities / subpopSize;
140 dl 1.2 }
141     else if (option.equals("-p"))
142 dl 1.4 subpopSize = Integer.parseInt(args[argc]);
143 dl 1.1 else if (option.equals("-g"))
144 dl 1.2 nGen = Integer.parseInt(args[argc]);
145     else if (option.equals("-n"))
146 dl 1.4 nSubpops = Integer.parseInt(args[argc]);
147     else if (option.equals("-q")) {
148     verbose = false;
149     argc--;
150     }
151     else if (option.equals("-r"))
152     nReps = Integer.parseInt(args[argc]);
153 dl 1.2 else
154 dl 1.1 maxThreads = Integer.parseInt(option);
155     argc++;
156     }
157     }
158     catch (Exception e) {
159     reportUsageErrorAndDie();
160     }
161    
162 dl 1.2 System.out.print("TSPExchangerTest");
163     System.out.print(" -c " + nCities);
164     System.out.print(" -g " + nGen);
165 dl 1.4 System.out.print(" -p " + subpopSize);
166     System.out.print(" -n " + nSubpops);
167     System.out.print(" -r " + nReps);
168 dl 1.2 System.out.print(" max threads " + maxThreads);
169     System.out.println();
170    
171     cities = new CitySet(nCities);
172 dl 1.1
173 dl 1.4 if (false && NCPUS > 4) {
174     int h = NCPUS/2;
175     System.out.printf("Threads: %4d Warmup\n", h);
176     oneRun(h, nSubpops, subpopSize, nGen);
177     Thread.sleep(500);
178     }
179    
180     int maxt = (maxThreads < nSubpops) ? maxThreads : nSubpops;
181     for (int j = 0; j < nReps; ++j) {
182     for (int i = 2; i <= maxt; i += 2) {
183     System.out.printf("Threads: %4d Replication: %2d\n", i, j);
184     oneRun(i, nSubpops, subpopSize, nGen);
185     Thread.sleep(500);
186     }
187     }
188 dl 1.1 }
189    
190 dl 1.2 static void reportUsageErrorAndDie() {
191     System.out.print("usage: TSPExchangerTest");
192     System.out.print(" [-c #cities]");
193 dl 1.4 System.out.print(" [-p #subpopSize]");
194 dl 1.2 System.out.print(" [-g #generations]");
195 dl 1.4 System.out.print(" [-n #subpops]");
196     System.out.print(" [-r #replications]");
197     System.out.print(" [-q <quiet>]");
198 dl 1.2 System.out.print(" #threads]");
199     System.out.println();
200 dl 1.1 System.exit(0);
201     }
202    
203 dl 1.2 /**
204 dl 1.4 * Perform one run with the given parameters. Each run complete
205     * when there are fewer than 2 active threads. When there is
206     * only one remaining thread, it will have no one to exchange
207     * with, so it is terminated (via interrupt).
208 dl 1.2 */
209 jsr166 1.6 static void oneRun(int nThreads, int nSubpops, int subpopSize, int nGen)
210 dl 1.2 throws InterruptedException {
211 dl 1.4 Population p = new Population(nThreads, nSubpops, subpopSize, nGen);
212 dl 1.2 ProgressMonitor mon = null;
213     if (verbose) {
214 dl 1.4 p.printSnapshot(0);
215 dl 1.2 mon = new ProgressMonitor(p);
216     mon.start();
217     }
218     long startTime = System.nanoTime();
219     p.start();
220 dl 1.4 p.awaitDone();
221 dl 1.2 long stopTime = System.nanoTime();
222     if (mon != null)
223     mon.interrupt();
224     p.shutdown();
225 dl 1.4 // Thread.sleep(100);
226 dl 1.2
227 dl 1.1 long elapsed = stopTime - startTime;
228 dl 1.2 double secs = (double)elapsed / 1000000000.0;
229     p.printSnapshot(secs);
230     }
231 dl 1.1
232    
233 dl 1.2 /**
234 dl 1.4 * A Population creates the subpops, subpops, and threads for a run
235 dl 1.2 * and has control methods to start, stop, and report progress.
236     */
237 dl 1.1 static final class Population {
238 dl 1.4 final Worker[] threads;
239     final Subpop[] subpops;
240 dl 1.2 final Exchanger<Strand> exchanger;
241     final CountDownLatch done;
242     final int nGen;
243 dl 1.4 final int subpopSize;
244 dl 1.1 final int nThreads;
245    
246 dl 1.4 Population(int nThreads, int nSubpops, int subpopSize, int nGen) {
247 dl 1.1 this.nThreads = nThreads;
248 dl 1.2 this.nGen = nGen;
249 dl 1.4 this.subpopSize = subpopSize;
250 dl 1.2 this.exchanger = new Exchanger<Strand>();
251 dl 1.4 this.done = new CountDownLatch(nThreads - 1);
252    
253     this.subpops = new Subpop[nSubpops];
254     for (int i = 0; i < nSubpops; i++)
255     subpops[i] = new Subpop(this);
256    
257     this.threads = new Worker[nThreads];
258     int maxExchanges = nGen * nSubpops / nThreads;
259     for (int i = 0; i < nThreads; ++i) {
260     threads[i] = new Worker(this, maxExchanges);
261     }
262    
263 dl 1.2 }
264    
265     void start() {
266 dl 1.4 for (int i = 0; i < nThreads; ++i) {
267 jsr166 1.6 threads[i].start();
268 dl 1.4 }
269 dl 1.2 }
270    
271     /** Stop the tasks */
272     void shutdown() {
273 dl 1.4 for (int i = 0; i < threads.length; ++ i)
274     threads[i].interrupt();
275 dl 1.2 }
276    
277 dl 1.4 void threadDone() {
278 dl 1.2 done.countDown();
279     }
280    
281 dl 1.4 /** Wait for tasks to complete */
282     void awaitDone() throws InterruptedException {
283 dl 1.2 done.await();
284     }
285    
286 dl 1.4 int totalExchanges() {
287     int xs = 0;
288 jsr166 1.6 for (int i = 0; i < threads.length; ++i)
289 dl 1.4 xs += threads[i].exchanges;
290     return xs;
291 dl 1.2 }
292    
293 dl 1.4 /**
294     * Prints statistics, including best and worst tour lengths
295     * for points scaled in [0,1), scaled by the square root of
296     * number of points. This simplifies checking results. The
297     * expected optimal TSP for random points is believed to be
298     * around 0.76 * sqrt(N). For papers discussing this, see
299     * http://www.densis.fee.unicamp.br/~moscato/TSPBIB_home.html
300     */
301 dl 1.2 void printSnapshot(double secs) {
302 dl 1.4 int xs = totalExchanges();
303     long rate = (xs == 0)? 0L : (long)((secs * 1000000000.0) / xs);
304     Chromosome bestc = subpops[0].chromosomes[0];
305 dl 1.3 Chromosome worstc = bestc;
306 dl 1.4 for (int k = 0; k < subpops.length; ++k) {
307     Chromosome[] cs = subpops[k].chromosomes;
308 dl 1.3 if (cs[0].fitness < bestc.fitness)
309     bestc = cs[0];
310     int w = cs[cs.length-1].fitness;
311     if (cs[cs.length-1].fitness > worstc.fitness)
312     worstc = cs[cs.length-1];
313     }
314     double sqrtn = Math.sqrt(cities.length);
315     double best = bestc.unitTourLength() / sqrtn;
316     double worst = worstc.unitTourLength() / sqrtn;
317 dl 1.4 System.out.printf("N:%4d T:%8.3f B:%6.3f W:%6.3f X:%9d R:%7d\n",
318     nThreads, secs, best, worst, xs, rate);
319     // exchanger.printStats();
320     // System.out.print(" s: " + exchanger.aveSpins());
321     // System.out.print(" p: " + exchanger.aveParks());
322 dl 1.2 }
323     }
324 dl 1.1
325 dl 1.2 /**
326 dl 1.4 * Worker threads perform updates on subpops.
327 dl 1.2 */
328 dl 1.4 static final class Worker extends Thread {
329     final Population pop;
330     final int maxExchanges;
331     int exchanges;
332     final RNG rng = new RNG();
333    
334     Worker(Population pop, int maxExchanges) {
335     this.pop = pop;
336     this.maxExchanges = maxExchanges;
337     }
338    
339     /**
340     * Repeatedly, find a subpop that is not being updated by
341     * another thread, and run a random number of updates on it.
342     */
343     public void run() {
344     try {
345     int len = pop.subpops.length;
346     int pos = (rng.next() & 0x7FFFFFFF) % len;
347     while (exchanges < maxExchanges) {
348     Subpop s = pop.subpops[pos];
349     AtomicBoolean busy = s.busy;
350     if (!busy.get() && busy.compareAndSet(false, true)) {
351     exchanges += s.runUpdates();
352     busy.set(false);
353     pos = (rng.next() & 0x7FFFFFFF) % len;
354     }
355     else if (++pos >= len)
356     pos = 0;
357     }
358     pop.threadDone();
359     } catch (InterruptedException fallthrough) {
360     }
361     }
362     }
363    
364     /**
365 jsr166 1.6 * A Subpop maintains a set of chromosomes..
366 dl 1.4 */
367     static final class Subpop {
368     /** The chromosomes, kept in sorted order */
369 dl 1.2 final Chromosome[] chromosomes;
370 dl 1.4 /** The parent population */
371 dl 1.2 final Population pop;
372 dl 1.4 /** Reservation bit for worker threads */
373     final AtomicBoolean busy;
374     /** The common exchanger, same for all subpops */
375 dl 1.2 final Exchanger<Strand> exchanger;
376     /** The current strand being exchanged */
377     Strand strand;
378     /** Bitset used in cross */
379     final int[] inTour;
380     final RNG rng;
381 dl 1.4 final int subpopSize;
382 dl 1.1
383 dl 1.4 Subpop(Population pop) {
384 dl 1.2 this.pop = pop;
385 dl 1.4 this.subpopSize = pop.subpopSize;
386 dl 1.2 this.exchanger = pop.exchanger;
387 dl 1.4 this.busy = new AtomicBoolean(false);
388 dl 1.2 this.rng = new RNG();
389     int length = cities.length;
390     this.strand = new Strand(length);
391     this.inTour = new int[(length >>> 5) + 1];
392 dl 1.4 this.chromosomes = new Chromosome[subpopSize];
393     for (int j = 0; j < subpopSize; ++j)
394 dl 1.2 chromosomes[j] = new Chromosome(length, rng);
395     Arrays.sort(chromosomes);
396     }
397    
398     /**
399 dl 1.4 * Run a random number of updates. The number of updates is
400     * at least 1 and no more than subpopSize. This
401     * controls the granularity of multiplexing subpop updates on
402     * to threads. It is small enough to balance out updates
403     * across tasks, but large enough to avoid having runs
404     * dominated by subpop selection. It is randomized to avoid
405     * long runs where pairs of subpops exchange only with each
406     * other. It is hardwired because small variations of it
407     * don't matter much.
408     *
409 jsr166 1.6 * @param g the first generation to run.
410 dl 1.4 */
411     int runUpdates() throws InterruptedException {
412     int n = 1 + (rng.next() & ((subpopSize << 1) - 1));
413     for (int i = 0; i < n; ++i)
414     update();
415     return n;
416 dl 1.1 }
417    
418 dl 1.2 /**
419 dl 1.4 * Choose a breeder, exchange strand with another subpop, and
420 dl 1.2 * cross them to create new chromosome to replace a chosen
421     * dyer.
422     */
423     void update() throws InterruptedException {
424     int b = chooseBreeder();
425     int d = chooseDyer(b);
426     Chromosome breeder = chromosomes[b];
427     Chromosome child = chromosomes[d];
428     chooseStrand(breeder);
429     strand = exchanger.exchange(strand);
430     cross(breeder, child);
431     fixOrder(child, d);
432     }
433    
434     /**
435     * Choose a breeder, with exponentially decreasing probability
436     * starting at best.
437     * @return index of selected breeder
438     */
439     int chooseBreeder() {
440 dl 1.4 int mask = (subpopSize >>> BREEDER_DECAY) - 1;
441 dl 1.2 int b = 0;
442     while ((rng.next() & mask) != mask) {
443 dl 1.4 if (++b >= subpopSize)
444 dl 1.2 b = 0;
445     }
446     return b;
447     }
448    
449     /**
450     * Choose a chromosome that will be replaced, with
451 jsr166 1.5 * exponentially decreasing probability starting at
452 dl 1.2 * worst, ignoring the excluded index
453     * @param exclude index to ignore; use -1 to not exclude any
454     * @return index of selected dyer
455     */
456     int chooseDyer(int exclude) {
457 dl 1.4 int mask = (subpopSize >>> DYER_DECAY) - 1;
458     int d = subpopSize - 1;
459 dl 1.2 while (d == exclude || (rng.next() & mask) != mask) {
460     if (--d < 0)
461 dl 1.4 d = subpopSize - 1;
462 dl 1.2 }
463     return d;
464     }
465    
466 jsr166 1.6 /**
467 dl 1.2 * Select a random strand of b's.
468     * @param breeder the breeder
469     */
470     void chooseStrand(Chromosome breeder) {
471     int[] bs = breeder.alleles;
472     int length = bs.length;
473     int strandLength = MIN_STRAND_LENGTH;
474     while (strandLength < length &&
475     (rng.next() & RANDOM_STRAND_MASK) != RANDOM_STRAND_MASK)
476     strandLength++;
477     strand.strandLength = strandLength;
478     int[] ss = strand.alleles;
479     int k = (rng.next() & 0x7FFFFFFF) % length;
480     for (int i = 0; i < strandLength; ++i) {
481     ss[i] = bs[k];
482     if (++k >= length) k = 0;
483     }
484     }
485    
486     /**
487     * Copy current strand to start of c's, and then append all
488     * remaining b's that aren't in the strand.
489     * @param breeder the breeder
490     * @param child the child
491     */
492     void cross(Chromosome breeder, Chromosome child) {
493     for (int k = 0; k < inTour.length; ++k) // clear bitset
494     inTour[k] = 0;
495    
496     // Copy current strand to c
497     int[] cs = child.alleles;
498     int ssize = strand.strandLength;
499     int[] ss = strand.alleles;
500     int i;
501     for (i = 0; i < ssize; ++i) {
502     int x = ss[i];
503     cs[i] = x;
504     inTour[x >>> 5] |= 1 << (x & 31); // record in bit set
505     }
506    
507     // Find index of matching origin in b
508     int first = cs[0];
509     int j = 0;
510     int[] bs = breeder.alleles;
511 jsr166 1.6 while (bs[j] != first)
512 dl 1.2 ++j;
513    
514     // Append remaining b's that aren't already in tour
515     while (i < cs.length) {
516     if (++j >= bs.length) j = 0;
517     int x = bs[j];
518     if ((inTour[x >>> 5] & (1 << (x & 31))) == 0)
519     cs[i++] = x;
520 jsr166 1.6 }
521 dl 1.1
522     }
523    
524 dl 1.2 /**
525     * Fix the sort order of a changed Chromosome c at position k
526     * @param c the chromosome
527 jsr166 1.6 * @param k the index
528 dl 1.2 */
529     void fixOrder(Chromosome c, int k) {
530     Chromosome[] cs = chromosomes;
531 dl 1.3 int oldFitness = c.fitness;
532 dl 1.2 c.recalcFitness();
533 dl 1.3 int newFitness = c.fitness;
534 dl 1.2 if (newFitness < oldFitness) {
535     int j = k;
536     int p = j - 1;
537     while (p >= 0 && cs[p].fitness > newFitness) {
538     cs[j] = cs[p];
539     j = p--;
540 dl 1.1 }
541 dl 1.2 cs[j] = c;
542     } else if (newFitness > oldFitness) {
543     int j = k;
544     int n = j + 1;
545     while (n < cs.length && cs[n].fitness < newFitness) {
546     cs[j] = cs[n];
547     j = n++;
548 dl 1.1 }
549 dl 1.2 cs[j] = c;
550 dl 1.1 }
551     }
552     }
553    
554 dl 1.2 /**
555     * A Chromosome is a candidate TSP tour.
556     */
557     static final class Chromosome implements Comparable {
558     /** Index of cities in tour order */
559     final int[] alleles;
560     /** Total tour length */
561 dl 1.3 int fitness;
562 dl 1.2
563 jsr166 1.6 /**
564 dl 1.2 * Initialize to random tour
565     */
566     Chromosome(int length, RNG random) {
567 dl 1.1 alleles = new int[length];
568     for (int i = 0; i < length; i++)
569     alleles[i] = i;
570     for (int i = length - 1; i > 0; i--) {
571 dl 1.2 int idx = (random.next() & 0x7FFFFFFF) % alleles.length;
572 dl 1.1 int tmp = alleles[i];
573     alleles[i] = alleles[idx];
574     alleles[idx] = tmp;
575     }
576     recalcFitness();
577     }
578 dl 1.2
579     public int compareTo(Object x) { // to enable sorting
580 dl 1.3 int xf = ((Chromosome)x).fitness;
581     int f = fitness;
582 dl 1.2 return ((f == xf)? 0 :((f < xf)? -1 : 1));
583 dl 1.1 }
584 dl 1.2
585 dl 1.1 void recalcFitness() {
586 dl 1.2 int[] a = alleles;
587     int len = a.length;
588     int p = a[0];
589 dl 1.3 long f = cities.distanceBetween(a[len-1], p);
590 dl 1.2 for (int i = 1; i < len; i++) {
591     int n = a[i];
592     f += cities.distanceBetween(p, n);
593     p = n;
594     }
595 dl 1.3 fitness = (int)(f / len);
596     }
597    
598 dl 1.4 /**
599     * Return tour length for points scaled in [0, 1).
600     */
601 dl 1.3 double unitTourLength() {
602     int[] a = alleles;
603     int len = a.length;
604     int p = a[0];
605     double f = cities.unitDistanceBetween(a[len-1], p);
606     for (int i = 1; i < len; i++) {
607     int n = a[i];
608     f += cities.unitDistanceBetween(p, n);
609     p = n;
610     }
611     return f;
612 dl 1.2 }
613    
614 dl 1.4 /**
615     * Check that this tour visits each city
616     */
617 jsr166 1.6 void validate() {
618 dl 1.2 int len = alleles.length;
619     boolean[] used = new boolean[len];
620 jsr166 1.6 for (int i = 0; i < len; ++i)
621 dl 1.2 used[alleles[i]] = true;
622 jsr166 1.6 for (int i = 0; i < len; ++i)
623 dl 1.2 if (!used[i])
624     throw new Error("Bad tour");
625 dl 1.1 }
626 dl 1.2
627     }
628    
629     /**
630 dl 1.4 * A Strand is a random sub-sequence of a Chromosome. Each subpop
631 dl 1.2 * creates only one strand, and then trades it with others,
632     * refilling it on each iteration.
633     */
634     static final class Strand {
635     final int[] alleles;
636     int strandLength;
637     Strand(int length) { alleles = new int[length]; }
638 dl 1.1 }
639 dl 1.2
640 dl 1.1 /**
641 jsr166 1.6 * A collection of (x,y) points that represent cities.
642 dl 1.1 */
643     static final class CitySet {
644 dl 1.2
645 dl 1.1 final int length;
646 dl 1.2 final int[] xPts;
647     final int[] yPts;
648 dl 1.3 final int[][] distances;
649 dl 1.2
650     CitySet(int n) {
651 dl 1.1 this.length = n;
652 dl 1.2 this.xPts = new int[n];
653     this.yPts = new int[n];
654 dl 1.3 this.distances = new int[n][n];
655 dl 1.2
656     RNG random = new RNG();
657 dl 1.1 for (int i = 0; i < n; i++) {
658 dl 1.2 xPts[i] = (random.next() & 0x7FFFFFFF);
659     yPts[i] = (random.next() & 0x7FFFFFFF);
660 dl 1.1 }
661 dl 1.2
662 dl 1.1 for (int i = 0; i < n; i++) {
663     for (int j = 0; j < n; j++) {
664 dl 1.3 double dx = (double)xPts[i] - (double)xPts[j];
665     double dy = (double)yPts[i] - (double)yPts[j];
666     double dd = Math.hypot(dx, dy) / 2.0;
667     long ld = Math.round(dd);
668 jsr166 1.6 distances[i][j] = (ld >= Integer.MAX_VALUE)?
669 dl 1.3 Integer.MAX_VALUE : (int)ld;
670 dl 1.1 }
671     }
672     }
673 dl 1.2
674 dl 1.3 /**
675     * Returns the cached distance between a pair of cities
676     */
677     int distanceBetween(int i, int j) {
678 dl 1.2 return distances[i][j];
679     }
680 dl 1.3
681     // Scale ints to doubles in [0,1)
682     static final double PSCALE = (double)0x80000000L;
683    
684     /**
685     * Return distance for points scaled in [0,1). This simplifies
686     * checking results. The expected optimal TSP for random
687     * points is believed to be around 0.76 * sqrt(N). For papers
688     * discussing this, see
689     * http://www.densis.fee.unicamp.br/~moscato/TSPBIB_home.html
690     */
691     double unitDistanceBetween(int i, int j) {
692     double dx = ((double)xPts[i] - (double)xPts[j]) / PSCALE;
693     double dy = ((double)yPts[i] - (double)yPts[j]) / PSCALE;
694     return Math.hypot(dx, dy);
695     }
696 jsr166 1.6
697 dl 1.2 }
698    
699     /**
700     * Cheap XorShift random number generator
701     */
702     static final class RNG {
703     /** Seed generator for XorShift RNGs */
704     static final Random seedGenerator = new Random();
705    
706     int seed;
707     RNG(int seed) { this.seed = seed; }
708 dl 1.3 RNG() { this.seed = seedGenerator.nextInt() | 1; }
709 dl 1.2
710     int next() {
711     int x = seed;
712     x ^= x << 6;
713     x ^= x >>> 21;
714     x ^= x << 7;
715     seed = x;
716     return x;
717 dl 1.1 }
718     }
719    
720 dl 1.2 static final class ProgressMonitor extends Thread {
721 dl 1.1 final Population pop;
722 dl 1.2 ProgressMonitor(Population p) { pop = p; }
723 dl 1.1 public void run() {
724 dl 1.2 double time = 0;
725 dl 1.1 try {
726 dl 1.2 while (!Thread.interrupted()) {
727     sleep(SNAPSHOT_RATE);
728     time += SNAPSHOT_RATE;
729     pop.printSnapshot(time / 1000.0);
730 dl 1.1 }
731 dl 1.2 } catch (InterruptedException ie) {}
732 dl 1.1 }
733     }
734     }