uct/uct.c

   1 #include <assert.h>
   2 #include <math.h>
   3 #include <stdio.h>
   4 #include <stdlib.h>
   5 #include <string.h>
   6 #include <time.h>
   7
   8 #define DEBUG
   9
  10 #include "debug.h"
  11 #include "board.h"
  12 #include "gtp.h"
  13 #include "move.h"
  14 #include "mq.h"
  15 #include "joseki/base.h"
  16 #include "playout.h"
  17 #include "playout/moggy.h"
  18 #include "playout/light.h"
  19 #include "tactics/util.h"
  20 #include "timeinfo.h"
  21 #include "uct/dynkomi.h"
  22 #include "uct/internal.h"
  23 #include "uct/plugins.h"
  24 #include "uct/prior.h"
  25 #include "uct/search.h"
  26 #include "uct/slave.h"
  27 #include "uct/tree.h"
  28 #include "uct/uct.h"
  29 #include "uct/walk.h"
  30
  31 struct uct_policy *policy_ucb1_init(struct uct *u, char *arg);
  32 struct uct_policy *policy_ucb1amaf_init(struct uct *u, char *arg, struct board *board);
  33 static void uct_pondering_start(struct uct *u, struct board *b0, struct tree *t, enum stone color);
  34
  35 /* Maximal simulation length. */
  36 #define MC_GAMELEN      MAX_GAMELEN
  37
  38
  39 static void
  40 setup_state(struct uct *u, struct board *b, enum stone color)
  41 {
  42         u->t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
  43                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, u->stats_hbits);
  44         if (u->force_seed)
  45                 fast_srandom(u->force_seed);
  46         if (UDEBUGL(3))
  47                 fprintf(stderr, "Fresh board with random seed %lu\n", fast_getseed());
  48         if (!u->no_tbook && b->moves == 0) {
  49                 assert(color == S_BLACK);
  50                 tree_load(u->t, b);
  51         }
  52 }
  53
  54 static void
  55 reset_state(struct uct *u)
  56 {
  57         assert(u->t);
  58         tree_done(u->t); u->t = NULL;
  59 }
  60
  61 static void
  62 setup_dynkomi(struct uct *u, struct board *b, enum stone to_play)
  63 {
  64         if (u->t->use_extra_komi && !u->pondering && u->dynkomi->permove)
  65                 u->t->extra_komi = u->dynkomi->permove(u->dynkomi, b, u->t);
  66         else if (!u->t->use_extra_komi)
  67                 u->t->extra_komi = 0;
  68 }
  69
  70 void
  71 uct_prepare_move(struct uct *u, struct board *b, enum stone color)
  72 {
  73         if (u->t) {
  74                 /* Verify that we have sane state. */
  75                 assert(b->es == u);
  76                 assert(u->t && b->moves);
  77                 if (color != stone_other(u->t->root_color)) {
  78                         fprintf(stderr, "Fatal: Non-alternating play detected %d %d\n",
  79                                 color, u->t->root_color);
  80                         exit(1);
  81                 }
  82                 uct_htable_reset(u->t);
  83
  84         } else {
  85                 /* We need fresh state. */
  86                 b->es = u;
  87                 setup_state(u, b, color);
  88         }
  89
  90         u->ownermap.playouts = 0;
  91         memset(u->ownermap.map, 0, board_size2(b) * sizeof(u->ownermap.map[0]));
  92         u->played_own = u->played_all = 0;
  93 }
  94
  95 static void
  96 dead_group_list(struct uct *u, struct board *b, struct move_queue *mq)
  97 {
  98         enum gj_state gs_array[board_size2(b)];
  99         struct group_judgement gj = { .thres = GJ_THRES, .gs = gs_array };
 100         board_ownermap_judge_groups(b, &u->ownermap, &gj);
 101         groups_of_status(b, &gj, GS_DEAD, mq);
 102 }
 103
 104 bool
 105 uct_pass_is_safe(struct uct *u, struct board *b, enum stone color, bool pass_all_alive)
 106 {
 107         /* Make sure enough playouts are simulated to get a reasonable dead group list. */
 108         while (u->ownermap.playouts < GJ_MINGAMES)
 109                 uct_playout(u, b, color, u->t);
 110
 111         struct move_queue mq = { .moves = 0 };
 112         dead_group_list(u, b, &mq);
 113         if (pass_all_alive && mq.moves > 0)
 114                 return false; // We need to remove some dead groups first.
 115         return pass_is_safe(b, color, &mq);
 116 }
 117
 118 static char *
 119 uct_printhook_ownermap(struct board *board, coord_t c, char *s, char *end)
 120 {
 121         struct uct *u = board->es;
 122         if (!u) {
 123                 strcat(s, ". ");
 124                 return s + 2;
 125         }
 126         const char chr[] = ":XO,"; // dame, black, white, unclear
 127         const char chm[] = ":xo,";
 128         char ch = chr[board_ownermap_judge_point(&u->ownermap, c, GJ_THRES)];
 129         if (ch == ',') { // less precise estimate then?
 130                 ch = chm[board_ownermap_judge_point(&u->ownermap, c, 0.67)];
 131         }
 132         s += snprintf(s, end - s, "%c ", ch);
 133         return s;
 134 }
 135
 136 static char *
 137 uct_notify_play(struct engine *e, struct board *b, struct move *m, char *enginearg)
 138 {
 139         struct uct *u = e->data;
 140         if (!u->t) {
 141                 /* No state, create one - this is probably game beginning
 142                  * and we need to load the opening tbook right now. */
 143                 uct_prepare_move(u, b, m->color);
 144                 assert(u->t);
 145         }
 146
 147         /* Stop pondering, required by tree_promote_at() */
 148         uct_pondering_stop(u);
 149         if (UDEBUGL(2) && u->slave)
 150                 tree_dump(u->t, u->dumpthres);
 151
 152         if (is_resign(m->coord)) {
 153                 /* Reset state. */
 154                 reset_state(u);
 155                 return NULL;
 156         }
 157
 158         /* Promote node of the appropriate move to the tree root. */
 159         assert(u->t->root);
 160         if (!tree_promote_at(u->t, b, m->coord)) {
 161                 if (UDEBUGL(3))
 162                         fprintf(stderr, "Warning: Cannot promote move node! Several play commands in row?\n");
 163                 reset_state(u);
 164                 return NULL;
 165         }
 166
 167         /* If we are a slave in a distributed engine, start pondering once
 168          * we know which move we actually played. See uct_genmove() about
 169          * the check for pass. */
 170         if (u->pondering_opt && u->slave && m->color == u->my_color && !is_pass(m->coord))
 171                 uct_pondering_start(u, b, u->t, stone_other(m->color));
 172
 173         return NULL;
 174 }
 175
 176 static char *
 177 uct_undo(struct engine *e, struct board *b)
 178 {
 179         struct uct *u = e->data;
 180
 181         if (!u->t) return NULL;
 182         uct_pondering_stop(u);
 183         reset_state(u);
 184         return NULL;
 185 }
 186
 187 static char *
 188 uct_result(struct engine *e, struct board *b)
 189 {
 190         struct uct *u = e->data;
 191         static char reply[1024];
 192
 193         if (!u->t)
 194                 return NULL;
 195         enum stone color = u->t->root_color;
 196         struct tree_node *n = u->t->root;
 197         snprintf(reply, 1024, "%s %s %d %.2f %.1f",
 198                  stone2str(color), coord2sstr(node_coord(n), b),
 199                  n->u.playouts, tree_node_get_value(u->t, -1, n->u.value),
 200                  u->t->use_extra_komi ? u->t->extra_komi : 0);
 201         return reply;
 202 }
 203
 204 static char *
 205 uct_chat(struct engine *e, struct board *b, char *cmd)
 206 {
 207         struct uct *u = e->data;
 208         static char reply[1024];
 209
 210         cmd += strspn(cmd, " \n\t");
 211         if (!strncasecmp(cmd, "winrate", 7)) {
 212                 if (!u->t)
 213                         return "no game context (yet?)";
 214                 enum stone color = u->t->root_color;
 215                 struct tree_node *n = u->t->root;
 216                 snprintf(reply, 1024, "In %d playouts at %d threads, %s %s can win with %.2f%% probability",
 217                          n->u.playouts, u->threads, stone2str(color), coord2sstr(node_coord(n), b),
 218                          tree_node_get_value(u->t, -1, n->u.value) * 100);
 219                 if (u->t->use_extra_komi && abs(u->t->extra_komi) >= 0.5) {
 220                         sprintf(reply + strlen(reply), ", while self-imposing extra komi %.1f",
 221                                 u->t->extra_komi);
 222                 }
 223                 strcat(reply, ".");
 224                 return reply;
 225         }
 226         return NULL;
 227 }
 228
 229 static void
 230 uct_dead_group_list(struct engine *e, struct board *b, struct move_queue *mq)
 231 {
 232         struct uct *u = e->data;
 233
 234         /* This means the game is probably over, no use pondering on. */
 235         uct_pondering_stop(u);
 236
 237         if (u->pass_all_alive)
 238                 return; // no dead groups
 239
 240         bool mock_state = false;
 241
 242         if (!u->t) {
 243                 /* No state, but we cannot just back out - we might
 244                  * have passed earlier, only assuming some stones are
 245                  * dead, and then re-connected, only to lose counting
 246                  * when all stones are assumed alive. */
 247                 uct_prepare_move(u, b, S_BLACK); assert(u->t);
 248                 mock_state = true;
 249         }
 250         /* Make sure the ownermap is well-seeded. */
 251         while (u->ownermap.playouts < GJ_MINGAMES)
 252                 uct_playout(u, b, S_BLACK, u->t);
 253         /* Show the ownermap: */
 254         if (DEBUGL(2))
 255                 board_print_custom(b, stderr, uct_printhook_ownermap);
 256
 257         dead_group_list(u, b, mq);
 258
 259         if (mock_state) {
 260                 /* Clean up the mock state in case we will receive
 261                  * a genmove; we could get a non-alternating-move
 262                  * error from uct_prepare_move() in that case otherwise. */
 263                 reset_state(u);
 264         }
 265 }
 266
 267 static void
 268 playout_policy_done(struct playout_policy *p)
 269 {
 270         if (p->done) p->done(p);
 271         if (p->data) free(p->data);
 272         free(p);
 273 }
 274
 275 static void
 276 uct_done(struct engine *e)
 277 {
 278         /* This is called on engine reset, especially when clear_board
 279          * is received and new game should begin. */
 280         struct uct *u = e->data;
 281         uct_pondering_stop(u);
 282         if (u->t) reset_state(u);
 283         free(u->ownermap.map);
 284
 285         free(u->policy);
 286         free(u->random_policy);
 287         playout_policy_done(u->playout);
 288         uct_prior_done(u->prior);
 289         joseki_done(u->jdict);
 290         pluginset_done(u->plugins);
 291 }
 292
 293
 294
 295 /* Run time-limited MCTS search on foreground. */
 296 static int
 297 uct_search(struct uct *u, struct board *b, struct time_info *ti, enum stone color, struct tree *t)
 298 {
 299         struct uct_search_state s;
 300         uct_search_start(u, b, color, t, ti, &s);
 301         if (UDEBUGL(2) && s.base_playouts > 0)
 302                 fprintf(stderr, "<pre-simulated %d games>\n", s.base_playouts);
 303
 304         /* The search tree is ctx->t. This is currently == . It is important
 305          * to reference ctx->t directly since the
 306          * thread manager will swap the tree pointer asynchronously. */
 307
 308         /* Now, just periodically poll the search tree. */
 309         /* Note that in case of TD_GAMES, threads will not wait for
 310          * the uct_search_check_stop() signalization. */
 311         while (1) {
 312                 time_sleep(TREE_BUSYWAIT_INTERVAL);
 313                 /* TREE_BUSYWAIT_INTERVAL should never be less than desired time, or the
 314                  * time control is broken. But if it happens to be less, we still search
 315                  * at least 100ms otherwise the move is completely random. */
 316
 317                 int i = uct_search_games(&s);
 318                 /* Print notifications etc. */
 319                 uct_search_progress(u, b, color, t, ti, &s, i);
 320                 /* Check if we should stop the search. */
 321                 if (uct_search_check_stop(u, b, color, t, ti, &s, i))
 322                         break;
 323         }
 324
 325         struct uct_thread_ctx *ctx = uct_search_stop();
 326         if (UDEBUGL(2)) tree_dump(t, u->dumpthres);
 327         if (UDEBUGL(2))
 328                 fprintf(stderr, "(avg score %f/%d; dynkomi's %f/%d value %f/%d)\n",
 329                         t->avg_score.value, t->avg_score.playouts,
 330                         u->dynkomi->score.value, u->dynkomi->score.playouts,
 331                         u->dynkomi->value.value, u->dynkomi->value.playouts);
 332         uct_progress_status(u, t, color, ctx->games, true);
 333
 334         u->played_own += ctx->games;
 335         return ctx->games;
 336 }
 337
 338 /* Start pondering background with @color to play. */
 339 static void
 340 uct_pondering_start(struct uct *u, struct board *b0, struct tree *t, enum stone color)
 341 {
 342         if (UDEBUGL(1))
 343                 fprintf(stderr, "Starting to ponder with color %s\n", stone2str(stone_other(color)));
 344         u->pondering = true;
 345
 346         /* We need a local board copy to ponder upon. */
 347         struct board *b = malloc2(sizeof(*b)); board_copy(b, b0);
 348
 349         /* *b0 did not have the genmove'd move played yet. */
 350         struct move m = { node_coord(t->root), t->root_color };
 351         int res = board_play(b, &m);
 352         assert(res >= 0);
 353         setup_dynkomi(u, b, stone_other(m.color));
 354
 355         /* Start MCTS manager thread "headless". */
 356         static struct uct_search_state s;
 357         uct_search_start(u, b, color, t, NULL, &s);
 358 }
 359
 360 /* uct_search_stop() frontend for the pondering (non-genmove) mode, and
 361  * to stop the background search for a slave in the distributed engine. */
 362 void
 363 uct_pondering_stop(struct uct *u)
 364 {
 365         if (!thread_manager_running)
 366                 return;
 367
 368         /* Stop the thread manager. */
 369         struct uct_thread_ctx *ctx = uct_search_stop();
 370         if (UDEBUGL(1)) {
 371                 if (u->pondering) fprintf(stderr, "(pondering) ");
 372                 uct_progress_status(u, ctx->t, ctx->color, ctx->games, true);
 373         }
 374         if (u->pondering) {
 375                 free(ctx->b);
 376                 u->pondering = false;
 377         }
 378 }
 379
 380
 381 void
 382 uct_genmove_setup(struct uct *u, struct board *b, enum stone color)
 383 {
 384         if (b->superko_violation) {
 385                 fprintf(stderr, "!!! WARNING: SUPERKO VIOLATION OCCURED BEFORE THIS MOVE\n");
 386                 fprintf(stderr, "Maybe you play with situational instead of positional superko?\n");
 387                 fprintf(stderr, "I'm going to ignore the violation, but note that I may miss\n");
 388                 fprintf(stderr, "some moves valid under this ruleset because of this.\n");
 389                 b->superko_violation = false;
 390         }
 391
 392         uct_prepare_move(u, b, color);
 393
 394         assert(u->t);
 395         u->my_color = color;
 396
 397         /* How to decide whether to use dynkomi in this game? Since we use
 398          * pondering, it's not simple "who-to-play" matter. Decide based on
 399          * the last genmove issued. */
 400         u->t->use_extra_komi = !!(u->dynkomi_mask & color);
 401         setup_dynkomi(u, b, color);
 402
 403         if (b->rules == RULES_JAPANESE)
 404                 u->territory_scoring = true;
 405
 406         /* Make pessimistic assumption about komi for Japanese rules to
 407          * avoid losing by 0.5 when winning by 0.5 with Chinese rules.
 408          * The rules usually give the same winner if the integer part of komi
 409          * is odd so we adjust the komi only if it is even (for a board of
 410          * odd size). We are not trying  to get an exact evaluation for rare
 411          * cases of seki. For details see http://home.snafu.de/jasiek/parity.html */
 412         if (u->territory_scoring && (((int)floor(b->komi) + board_size(b)) & 1)) {
 413                 b->komi += (color == S_BLACK ? 1.0 : -1.0);
 414                 if (UDEBUGL(0))
 415                         fprintf(stderr, "Setting komi to %.1f assuming Japanese rules\n",
 416                                 b->komi);
 417         }
 418 }
 419
 420 static coord_t *
 421 uct_genmove(struct engine *e, struct board *b, struct time_info *ti, enum stone color, bool pass_all_alive)
 422 {
 423         double start_time = time_now();
 424         struct uct *u = e->data;
 425         u->pass_all_alive |= pass_all_alive;
 426         uct_pondering_stop(u);
 427         uct_genmove_setup(u, b, color);
 428
 429         /* Start the Monte Carlo Tree Search! */
 430         int base_playouts = u->t->root->u.playouts;
 431         int played_games = uct_search(u, b, ti, color, u->t);
 432
 433         coord_t best_coord;
 434         struct tree_node *best;
 435         best = uct_search_result(u, b, color, u->pass_all_alive, played_games, base_playouts, &best_coord);
 436
 437         if (UDEBUGL(2)) {
 438                 double time = time_now() - start_time + 0.000001; /* avoid divide by zero */
 439                 fprintf(stderr, "genmove in %0.2fs (%d games/s, %d games/s/thread)\n",
 440                         time, (int)(played_games/time), (int)(played_games/time/u->threads));
 441         }
 442
 443         if (!best) {
 444                 /* Pass or resign. */
 445                 reset_state(u);
 446                 return coord_copy(best_coord);
 447         }
 448         tree_promote_node(u->t, &best);
 449
 450         /* After a pass, pondering is harmful for two reasons:
 451          * (i) We might keep pondering even when the game is over.
 452          * Of course this is the case for opponent resign as well.
 453          * (ii) More importantly, the ownermap will get skewed since
 454          * the UCT will start cutting off any playouts. */
 455         if (u->pondering_opt && !is_pass(node_coord(best))) {
 456                 uct_pondering_start(u, b, u->t, stone_other(color));
 457         }
 458         return coord_copy(best_coord);
 459 }
 460
 461
 462 bool
 463 uct_gentbook(struct engine *e, struct board *b, struct time_info *ti, enum stone color)
 464 {
 465         struct uct *u = e->data;
 466         if (!u->t) uct_prepare_move(u, b, color);
 467         assert(u->t);
 468
 469         if (ti->dim == TD_GAMES) {
 470                 /* Don't count in games that already went into the tbook. */
 471                 ti->len.games += u->t->root->u.playouts;
 472         }
 473         uct_search(u, b, ti, color, u->t);
 474
 475         assert(ti->dim == TD_GAMES);
 476         tree_save(u->t, b, ti->len.games / 100);
 477
 478         return true;
 479 }
 480
 481 void
 482 uct_dumptbook(struct engine *e, struct board *b, enum stone color)
 483 {
 484         struct uct *u = e->data;
 485         struct tree *t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
 486                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, 0);
 487         tree_load(t, b);
 488         tree_dump(t, 0);
 489         tree_done(t);
 490 }
 491
 492
 493 floating_t
 494 uct_evaluate_one(struct engine *e, struct board *b, struct time_info *ti, coord_t c, enum stone color)
 495 {
 496         struct uct *u = e->data;
 497
 498         struct board b2;
 499         board_copy(&b2, b);
 500         struct move m = { c, color };
 501         int res = board_play(&b2, &m);
 502         if (res < 0)
 503                 return NAN;
 504         color = stone_other(color);
 505
 506         if (u->t) reset_state(u);
 507         uct_prepare_move(u, &b2, color);
 508         assert(u->t);
 509
 510         floating_t bestval;
 511         uct_search(u, &b2, ti, color, u->t);
 512         struct tree_node *best = u->policy->choose(u->policy, u->t->root, &b2, color, resign);
 513         if (!best) {
 514                 bestval = NAN; // the opponent has no reply!
 515         } else {
 516                 bestval = tree_node_get_value(u->t, 1, best->u.value);
 517         }
 518
 519         reset_state(u); // clean our junk
 520
 521         return isnan(bestval) ? NAN : 1.0f - bestval;
 522 }
 523
 524 void
 525 uct_evaluate(struct engine *e, struct board *b, struct time_info *ti, floating_t *vals, enum stone color)
 526 {
 527         for (int i = 0; i < b->flen; i++) {
 528                 if (is_pass(b->f[i]))
 529                         vals[i] = NAN;
 530                 else
 531                         vals[i] = uct_evaluate_one(e, b, ti, b->f[i], color);
 532         }
 533 }
 534
 535
 536 struct uct *
 537 uct_state_init(char *arg, struct board *b)
 538 {
 539         struct uct *u = calloc2(1, sizeof(struct uct));
 540         bool pat_setup = false;
 541
 542         char * spatial_dict_filename = default_spatial_dict_filename ;
 543
 544         u->debug_level = debug_level;
 545         u->reportfreq = 10000;
 546         u->gamelen = MC_GAMELEN;
 547         u->resign_threshold = 0.2;
 548         u->sure_win_threshold = 0.9;
 549         u->mercymin = 0;
 550         u->significant_threshold = 50;
 551         u->expand_p = 8;
 552         u->dumpthres = 1000;
 553         u->playout_amaf = true;
 554         u->amaf_prior = false;
 555         u->max_tree_size = 1408ULL * 1048576;
 556         u->fast_alloc = true;
 557         u->pruning_threshold = 0;
 558
 559         u->threads = 1;
 560         u->thread_model = TM_TREEVL;
 561         u->virtual_loss = 1;
 562
 563         u->fuseki_end = 20; // max time at 361*20% = 72 moves (our 36th move, still 99 to play)
 564         u->yose_start = 40; // (100-40-25)*361/100/2 = 63 moves still to play by us then
 565         u->bestr_ratio = 0.02;
 566         // 2.5 is clearly too much, but seems to compensate well for overly stern time allocations.
 567         // TODO: Further tuning and experiments with better time allocation schemes.
 568         u->best2_ratio = 2.5;
 569         u->max_maintime_ratio = 3.0;
 570
 571         u->val_scale = 0; u->val_points = 40;
 572         u->dynkomi_interval = 1000;
 573         u->dynkomi_mask = S_BLACK | S_WHITE;
 574
 575         u->tenuki_d = 4;
 576         u->local_tree_aging = 80;
 577         u->local_tree_depth_decay = 1.5;
 578         u->local_tree_eval = LTE_ROOT;
 579         u->local_tree_neival = true;
 580
 581         u->max_slaves = -1;
 582         u->slave_index = -1;
 583         u->stats_delay = 0.01; // 10 ms
 584
 585         u->plugins = pluginset_init(b);
 586
 587         u->jdict = joseki_load(b->size);
 588
 589         if (arg) {
 590                 char *optspec, *next = arg;
 591                 while (*next) {
 592                         optspec = next;
 593                         next += strcspn(next, ",");
 594                         if (*next) { *next++ = 0; } else { *next = 0; }
 595
 596                         char *optname = optspec;
 597                         char *optval = strchr(optspec, '=');
 598                         if (optval) *optval++ = 0;
 599
 600                         /** Basic options */
 601
 602                         if (!strcasecmp(optname, "debug")) {
 603                                 if (optval)
 604                                         u->debug_level = atoi(optval);
 605                                 else
 606                                         u->debug_level++;
 607                         } else if (!strcasecmp(optname, "reporting") && optval) {
 608                                 /* The format of output for detailed progress
 609                                  * information (such as current best move and
 610                                  * its value, etc.). */
 611                                 if (!strcasecmp(optval, "text")) {
 612                                         /* Plaintext traditional output. */
 613                                         u->reporting = UR_TEXT;
 614                                 } else if (!strcasecmp(optval, "json")) {
 615                                         /* JSON output. Implies debug=0. */
 616                                         u->reporting = UR_JSON;
 617                                         u->debug_level = 0;
 618                                 } else if (!strcasecmp(optval, "jsonbig")) {
 619                                         /* JSON output, but much more detailed.
 620                                          * Implies debug=0. */
 621                                         u->reporting = UR_JSON_BIG;
 622                                         u->debug_level = 0;
 623                                 } else {
 624                                         fprintf(stderr, "UCT: Invalid reporting format %s\n", optval);
 625                                         exit(1);
 626                                 }
 627                         } else if (!strcasecmp(optname, "reportfreq") && optval) {
 628                                 /* The progress information line will be shown
 629                                  * every <reportfreq> simulations. */
 630                                 u->reportfreq = atoi(optval);
 631                         } else if (!strcasecmp(optname, "dumpthres") && optval) {
 632                                 /* When dumping the UCT tree on output, include
 633                                  * nodes with at least this many playouts.
 634                                  * (This value is re-scaled "intelligently"
 635                                  * in case of very large trees.) */
 636                                 u->dumpthres = atoi(optval);
 637                         } else if (!strcasecmp(optname, "resign_threshold") && optval) {
 638                                 /* Resign when this ratio of games is lost
 639                                  * after GJ_MINGAMES sample is taken. */
 640                                 u->resign_threshold = atof(optval);
 641                         } else if (!strcasecmp(optname, "sure_win_threshold") && optval) {
 642                                 /* Stop reading when this ratio of games is won
 643                                  * after PLAYOUT_EARLY_BREAK_MIN sample is
 644                                  * taken. (Prevents stupid time losses,
 645                                  * friendly to human opponents.) */
 646                                 u->sure_win_threshold = atof(optval);
 647                         } else if (!strcasecmp(optname, "force_seed") && optval) {
 648                                 /* Set RNG seed at the tree setup. */
 649                                 u->force_seed = atoi(optval);
 650                         } else if (!strcasecmp(optname, "no_tbook")) {
 651                                 /* Disable UCT opening tbook. */
 652                                 u->no_tbook = true;
 653                         } else if (!strcasecmp(optname, "pass_all_alive")) {
 654                                 /* Whether to consider passing only after all
 655                                  * dead groups were removed from the board;
 656                                  * this is like all genmoves are in fact
 657                                  * kgs-genmove_cleanup. */
 658                                 u->pass_all_alive = !optval || atoi(optval);
 659                         } else if (!strcasecmp(optname, "territory_scoring")) {
 660                                 /* Use territory scoring (default is area scoring).
 661                                  * An explicit kgs-rules command overrides this. */
 662                                 u->territory_scoring = !optval || atoi(optval);
 663                         } else if (!strcasecmp(optname, "stones_only")) {
 664                                 /* Do not count eyes. Nice to teach go to kids.
 665                                  * http://strasbourg.jeudego.org/regle_strasbourgeoise.htm */
 666                                 b->rules = RULES_STONES_ONLY;
 667                                 u->pass_all_alive = true;
 668                         } else if (!strcasecmp(optname, "banner") && optval) {
 669                                 /* Additional banner string. This must come as the
 670                                  * last engine parameter. */
 671                                 if (*next) *--next = ',';
 672                                 u->banner = strdup(optval);
 673                                 break;
 674                         } else if (!strcasecmp(optname, "plugin") && optval) {
 675                                 /* Load an external plugin; filename goes before the colon,
 676                                  * extra arguments after the colon. */
 677                                 char *pluginarg = strchr(optval, ':');
 678                                 if (pluginarg)
 679                                         *pluginarg++ = 0;
 680                                 plugin_load(u->plugins, optval, pluginarg);
 681
 682                         /** UCT behavior and policies */
 683
 684                         } else if ((!strcasecmp(optname, "policy")
 685                                 /* Node selection policy. ucb1amaf is the
 686                                  * default policy implementing RAVE, while
 687                                  * ucb1 is the simple exploration/exploitation
 688                                  * policy. Policies can take further extra
 689                                  * options. */
 690                                     || !strcasecmp(optname, "random_policy")) && optval) {
 691                                 /* A policy to be used randomly with small
 692                                  * chance instead of the default policy. */
 693                                 char *policyarg = strchr(optval, ':');
 694                                 struct uct_policy **p = !strcasecmp(optname, "policy") ? &u->policy : &u->random_policy;
 695                                 if (policyarg)
 696                                         *policyarg++ = 0;
 697                                 if (!strcasecmp(optval, "ucb1")) {
 698                                         *p = policy_ucb1_init(u, policyarg);
 699                                 } else if (!strcasecmp(optval, "ucb1amaf")) {
 700                                         *p = policy_ucb1amaf_init(u, policyarg, b);
 701                                 } else {
 702                                         fprintf(stderr, "UCT: Invalid tree policy %s\n", optval);
 703                                         exit(1);
 704                                 }
 705                         } else if (!strcasecmp(optname, "playout") && optval) {
 706                                 /* Random simulation (playout) policy.
 707                                  * moggy is the default policy with large
 708                                  * amount of domain-specific knowledge and
 709                                  * heuristics. light is a simple uniformly
 710                                  * random move selection policy. */
 711                                 char *playoutarg = strchr(optval, ':');
 712                                 if (playoutarg)
 713                                         *playoutarg++ = 0;
 714                                 if (!strcasecmp(optval, "moggy")) {
 715                                         u->playout = playout_moggy_init(playoutarg, b, u->jdict);
 716                                 } else if (!strcasecmp(optval, "light")) {
 717                                         u->playout = playout_light_init(playoutarg, b);
 718                                 } else {
 719                                         fprintf(stderr, "UCT: Invalid playout policy %s\n", optval);
 720                                         exit(1);
 721                                 }
 722                         } else if (!strcasecmp(optname, "prior") && optval) {
 723                                 /* Node priors policy. When expanding a node,
 724                                  * it will seed node values heuristically
 725                                  * (most importantly, based on playout policy
 726                                  * opinion, but also with regard to other
 727                                  * things). See uct/prior.c for details.
 728                                  * Use prior=eqex=0 to disable priors. */
 729                                 u->prior = uct_prior_init(optval, b, u);
 730                         } else if (!strcasecmp(optname, "mercy") && optval) {
 731                                 /* Minimal difference of black/white captures
 732                                  * to stop playout - "Mercy Rule". Speeds up
 733                                  * hopeless playouts at the expense of some
 734                                  * accuracy. */
 735                                 u->mercymin = atoi(optval);
 736                         } else if (!strcasecmp(optname, "gamelen") && optval) {
 737                                 /* Maximum length of single simulation
 738                                  * in moves. */
 739                                 u->gamelen = atoi(optval);
 740                         } else if (!strcasecmp(optname, "expand_p") && optval) {
 741                                 /* Expand UCT nodes after it has been
 742                                  * visited this many times. */
 743                                 u->expand_p = atoi(optval);
 744                         } else if (!strcasecmp(optname, "random_policy_chance") && optval) {
 745                                 /* If specified (N), with probability 1/N, random_policy policy
 746                                  * descend is used instead of main policy descend; useful
 747                                  * if specified policy (e.g. UCB1AMAF) can make unduly biased
 748                                  * choices sometimes, you can fall back to e.g.
 749                                  * random_policy=UCB1. */
 750                                 u->random_policy_chance = atoi(optval);
 751
 752                         /** General AMAF behavior */
 753                         /* (Only relevant if the policy supports AMAF.
 754                          * More variables can be tuned as policy
 755                          * parameters.) */
 756
 757                         } else if (!strcasecmp(optname, "playout_amaf")) {
 758                                 /* Whether to include random playout moves in
 759                                  * AMAF as well. (Otherwise, only tree moves
 760                                  * are included in AMAF. Of course makes sense
 761                                  * only in connection with an AMAF policy.) */
 762                                 /* with-without: 55.5% (+-4.1) */
 763                                 if (optval && *optval == '0')
 764                                         u->playout_amaf = false;
 765                                 else
 766                                         u->playout_amaf = true;
 767                         } else if (!strcasecmp(optname, "playout_amaf_cutoff") && optval) {
 768                                 /* Keep only first N% of playout stage AMAF
 769                                  * information. */
 770                                 u->playout_amaf_cutoff = atoi(optval);
 771                         } else if (!strcasecmp(optname, "amaf_prior") && optval) {
 772                                 /* In node policy, consider prior values
 773                                  * part of the real result term or part
 774                                  * of the AMAF term? */
 775                                 u->amaf_prior = atoi(optval);
 776
 777                         /** Performance and memory management */
 778
 779                         } else if (!strcasecmp(optname, "threads") && optval) {
 780                                 /* By default, Pachi will run with only single
 781                                  * tree search thread! */
 782                                 u->threads = atoi(optval);
 783                         } else if (!strcasecmp(optname, "thread_model") && optval) {
 784                                 if (!strcasecmp(optval, "tree")) {
 785                                         /* Tree parallelization - all threads
 786                                          * grind on the same tree. */
 787                                         u->thread_model = TM_TREE;
 788                                         u->virtual_loss = 0;
 789                                 } else if (!strcasecmp(optval, "treevl")) {
 790                                         /* Tree parallelization, but also
 791                                          * with virtual losses - this discou-
 792                                          * rages most threads choosing the
 793                                          * same tree branches to read. */
 794                                         u->thread_model = TM_TREEVL;
 795                                 } else {
 796                                         fprintf(stderr, "UCT: Invalid thread model %s\n", optval);
 797                                         exit(1);
 798                                 }
 799                         } else if (!strcasecmp(optname, "virtual_loss")) {
 800                                 /* Number of virtual losses added before evaluating a node. */
 801                                 u->virtual_loss = !optval || atoi(optval);
 802                         } else if (!strcasecmp(optname, "pondering")) {
 803                                 /* Keep searching even during opponent's turn. */
 804                                 u->pondering_opt = !optval || atoi(optval);
 805                         } else if (!strcasecmp(optname, "max_tree_size") && optval) {
 806                                 /* Maximum amount of memory [MiB] consumed by the move tree.
 807                                  * For fast_alloc it includes the temp tree used for pruning.
 808                                  * Default is 3072 (3 GiB). */
 809                                 u->max_tree_size = atol(optval) * 1048576;
 810                         } else if (!strcasecmp(optname, "fast_alloc")) {
 811                                 u->fast_alloc = !optval || atoi(optval);
 812                         } else if (!strcasecmp(optname, "pruning_threshold") && optval) {
 813                                 /* Force pruning at beginning of a move if the tree consumes
 814                                  * more than this [MiB]. Default is 10% of max_tree_size.
 815                                  * Increase to reduce pruning time overhead if memory is plentiful.
 816                                  * This option is meaningful only for fast_alloc. */
 817                                 u->pruning_threshold = atol(optval) * 1048576;
 818
 819                         /** Time control */
 820
 821                         } else if (!strcasecmp(optname, "best2_ratio") && optval) {
 822                                 /* If set, prolong simulating while
 823                                  * first_best/second_best playouts ratio
 824                                  * is less than best2_ratio. */
 825                                 u->best2_ratio = atof(optval);
 826                         } else if (!strcasecmp(optname, "bestr_ratio") && optval) {
 827                                 /* If set, prolong simulating while
 828                                  * best,best_best_child values delta
 829                                  * is more than bestr_ratio. */
 830                                 u->bestr_ratio = atof(optval);
 831                         } else if (!strcasecmp(optname, "max_maintime_ratio") && optval) {
 832                                 /* If set and while not in byoyomi, prolong simulating no more than
 833                                  * max_maintime_ratio times the normal desired thinking time. */
 834                                 u->max_maintime_ratio = atof(optval);
 835                         } else if (!strcasecmp(optname, "fuseki_end") && optval) {
 836                                 /* At the very beginning it's not worth thinking
 837                                  * too long because the playout evaluations are
 838                                  * very noisy. So gradually increase the thinking
 839                                  * time up to maximum when fuseki_end percent
 840                                  * of the board has been played.
 841                                  * This only applies if we are not in byoyomi. */
 842                                 u->fuseki_end = atoi(optval);
 843                         } else if (!strcasecmp(optname, "yose_start") && optval) {
 844                                 /* When yose_start percent of the board has been
 845                                  * played, or if we are in byoyomi, stop spending
 846                                  * more time and spread the remaining time
 847                                  * uniformly.
 848                                  * Between fuseki_end and yose_start, we spend
 849                                  * a constant proportion of the remaining time
 850                                  * on each move. (yose_start should actually
 851                                  * be much earlier than when real yose start,
 852                                  * but "yose" is a good short name to convey
 853                                  * the idea.) */
 854                                 u->yose_start = atoi(optval);
 855
 856                         /** Dynamic komi */
 857
 858                         } else if (!strcasecmp(optname, "dynkomi") && optval) {
 859                                 /* Dynamic komi approach; there are multiple
 860                                  * ways to adjust komi dynamically throughout
 861                                  * play. We currently support two: */
 862                                 char *dynkomiarg = strchr(optval, ':');
 863                                 if (dynkomiarg)
 864                                         *dynkomiarg++ = 0;
 865                                 if (!strcasecmp(optval, "none")) {
 866                                         u->dynkomi = uct_dynkomi_init_none(u, dynkomiarg, b);
 867                                 } else if (!strcasecmp(optval, "linear")) {
 868                                         /* You should set dynkomi_mask=1 or a very low
 869                                          * handicap_value for white. */
 870                                         u->dynkomi = uct_dynkomi_init_linear(u, dynkomiarg, b);
 871                                 } else if (!strcasecmp(optval, "adaptive")) {
 872                                         /* There are many more knobs to
 873                                          * crank - see uct/dynkomi.c. */
 874                                         u->dynkomi = uct_dynkomi_init_adaptive(u, dynkomiarg, b);
 875                                 } else {
 876                                         fprintf(stderr, "UCT: Invalid dynkomi mode %s\n", optval);
 877                                         exit(1);
 878                                 }
 879                         } else if (!strcasecmp(optname, "dynkomi_mask") && optval) {
 880                                 /* Bitmask of colors the player must be
 881                                  * for dynkomi be applied; the default dynkomi_mask=3 allows
 882                                  * dynkomi even in games where Pachi is white. */
 883                                 u->dynkomi_mask = atoi(optval);
 884                         } else if (!strcasecmp(optname, "dynkomi_interval") && optval) {
 885                                 /* If non-zero, re-adjust dynamic komi
 886                                  * throughout a single genmove reading,
 887                                  * roughly every N simulations. */
 888                                 /* XXX: Does not work with tree
 889                                  * parallelization. */
 890                                 u->dynkomi_interval = atoi(optval);
 891
 892                         /** Node value result scaling */
 893
 894                         } else if (!strcasecmp(optname, "val_scale") && optval) {
 895                                 /* How much of the game result value should be
 896                                  * influenced by win size. Zero means it isn't. */
 897                                 u->val_scale = atof(optval);
 898                         } else if (!strcasecmp(optname, "val_points") && optval) {
 899                                 /* Maximum size of win to be scaled into game
 900                                  * result value. Zero means boardsize^2. */
 901                                 u->val_points = atoi(optval) * 2; // result values are doubled
 902                         } else if (!strcasecmp(optname, "val_extra")) {
 903                                 /* If false, the score coefficient will be simply
 904                                  * added to the value, instead of scaling the result
 905                                  * coefficient because of it. */
 906                                 u->val_extra = !optval || atoi(optval);
 907
 908                         /** Local trees */
 909                         /* (Purely experimental. Does not work - yet!) */
 910
 911                         } else if (!strcasecmp(optname, "local_tree")) {
 912                                 /* Whether to bias exploration by local tree values. */
 913                                 u->local_tree = !optval || atoi(optval);
 914                         } else if (!strcasecmp(optname, "tenuki_d") && optval) {
 915                                 /* Tenuki distance at which to break the local tree. */
 916                                 u->tenuki_d = atoi(optval);
 917                                 if (u->tenuki_d > TREE_NODE_D_MAX + 1) {
 918                                         fprintf(stderr, "uct: tenuki_d must not be larger than TREE_NODE_D_MAX+1 %d\n", TREE_NODE_D_MAX + 1);
 919                                         exit(1);
 920                                 }
 921                         } else if (!strcasecmp(optname, "local_tree_aging") && optval) {
 922                                 /* How much to reduce local tree values between moves. */
 923                                 u->local_tree_aging = atof(optval);
 924                         } else if (!strcasecmp(optname, "local_tree_depth_decay") && optval) {
 925                                 /* With value x>0, during the descent the node
 926                                  * contributes 1/x^depth playouts in
 927                                  * the local tree. I.e., with x>1, nodes more
 928                                  * distant from local situation contribute more
 929                                  * than nodes near the root. */
 930                                 u->local_tree_depth_decay = atof(optval);
 931                         } else if (!strcasecmp(optname, "local_tree_allseq")) {
 932                                 /* If disabled, only complete sequences are stored
 933                                  * in the local tree. If this is on, also
 934                                  * subsequences starting at each move are stored. */
 935                                 u->local_tree_allseq = !optval || atoi(optval);
 936                         } else if (!strcasecmp(optname, "local_tree_neival")) {
 937                                 /* If disabled, local node value is not
 938                                  * computed just based on terminal status
 939                                  * of the coordinate, but also its neighbors. */
 940                                 u->local_tree_neival = !optval || atoi(optval);
 941                         } else if (!strcasecmp(optname, "local_tree_eval")) {
 942                                 /* How is the value inserted in the local tree
 943                                  * determined. */
 944                                 if (!strcasecmp(optval, "root"))
 945                                         /* All moves within a tree branch are
 946                                          * considered wrt. their merit
 947                                          * reaching tachtical goal of making
 948                                          * the first move in the branch
 949                                          * survive. */
 950                                         u->local_tree_eval = LTE_ROOT;
 951                                 else if (!strcasecmp(optval, "each"))
 952                                         /* Each move is considered wrt.
 953                                          * its own survival. */
 954                                         u->local_tree_eval = LTE_EACH;
 955                                 else if (!strcasecmp(optval, "total"))
 956                                         /* The tactical goal is the survival
 957                                          * of all the moves of my color and
 958                                          * non-survival of all the opponent
 959                                          * moves. Local values (and their
 960                                          * inverses) are averaged. */
 961                                         u->local_tree_eval = LTE_TOTAL;
 962                                 else {
 963                                         fprintf(stderr, "uct: unknown local_tree_eval %s\n", optval);
 964                                         exit(1);
 965                                 }
 966                         } else if (!strcasecmp(optname, "local_tree_rootchoose")) {
 967                                 /* If disabled, only moves within the local
 968                                  * tree branch are considered; the values
 969                                  * of the branch roots (i.e. root children)
 970                                  * are ignored. This may make sense together
 971                                  * with eval!=each, we consider only moves
 972                                  * that influence the goal, not the "rating"
 973                                  * of the goal itself. (The real solution
 974                                  * will be probably using criticality to pick
 975                                  * local tree branches.) */
 976                                 u->local_tree_rootchoose = !optval || atoi(optval);
 977
 978                         /** Other heuristics */
 979                         } else if (!strcasecmp(optname, "patterns")) {
 980                                 /* Load pattern database. Various modules
 981                                  * (priors, policies etc.) may make use
 982                                  * of this database. They will request
 983                                  * it automatically in that case, but you
 984                                  * can use this option to tweak the pattern
 985                                  * parameters. */
 986                                 patterns_init(&u->pat, optval, false, true, spatial_dict_filename);
 987                                 u->want_pat = pat_setup = true;
 988                         } else if (!strcasecmp(optname, "significant_threshold") && optval) {
 989                                 /* Some heuristics (XXX: none in mainline) rely
 990                                  * on the knowledge of the last "significant"
 991                                  * node in the descent. Such a node is
 992                                  * considered reasonably trustworthy to carry
 993                                  * some meaningful information in the values
 994                                  * of the node and its children. */
 995                                 u->significant_threshold = atoi(optval);
 996
 997                         /** Distributed engine slaves setup */
 998
 999                         } else if (!strcasecmp(optname, "slave")) {
1000                                 /* Act as slave for the distributed engine. */
1001                                 u->slave = !optval || atoi(optval);
1002                         } else if (!strcasecmp(optname, "slave_index") && optval) {
1003                                 /* Optional index if per-slave behavior is desired.
1004                                  * Must be given as index/max */
1005                                 u->slave_index = atoi(optval);
1006                                 char *p = strchr(optval, '/');
1007                                 if (p) u->max_slaves = atoi(++p);
1008                         } else if (!strcasecmp(optname, "shared_nodes") && optval) {
1009                                 /* Share at most shared_nodes between master and slave at each genmoves.
1010                                  * Must use the same value in master and slaves. */
1011                                 u->shared_nodes = atoi(optval);
1012                         } else if (!strcasecmp(optname, "shared_levels") && optval) {
1013                                 /* Share only nodes of level <= shared_levels. */
1014                                 u->shared_levels = atoi(optval);
1015                         } else if (!strcasecmp(optname, "stats_hbits") && optval) {
1016                                 /* Set hash table size to 2^stats_hbits for the shared stats. */
1017                                 u->stats_hbits = atoi(optval);
1018                         } else if (!strcasecmp(optname, "stats_delay") && optval) {
1019                                 /* How long to wait in slave for initial stats to build up before
1020                                  * replying to the genmoves command (in ms) */
1021                                 u->stats_delay = 0.001 * atof(optval);
1022
1023                         } else {
1024                                 fprintf(stderr, "uct: Invalid engine argument %s or missing value\n", optname);
1025                                 exit(1);
1026                         }
1027                 }
1028         }
1029
1030         if (!u->policy)
1031                 u->policy = policy_ucb1amaf_init(u, NULL, b);
1032
1033         if (!!u->random_policy_chance ^ !!u->random_policy) {
1034                 fprintf(stderr, "uct: Only one of random_policy and random_policy_chance is set\n");
1035                 exit(1);
1036         }
1037
1038         if (!u->local_tree) {
1039                 /* No ltree aging. */
1040                 u->local_tree_aging = 1.0f;
1041         }
1042
1043         if (u->fast_alloc) {
1044                 if (u->pruning_threshold < u->max_tree_size / 10)
1045                         u->pruning_threshold = u->max_tree_size / 10;
1046                 if (u->pruning_threshold > u->max_tree_size / 2)
1047                         u->pruning_threshold = u->max_tree_size / 2;
1048
1049                 /* Limit pruning temp space to 20% of memory. Beyond this we discard
1050                  * the nodes and recompute them at the next move if necessary. */
1051                 u->max_pruned_size = u->max_tree_size / 5;
1052                 u->max_tree_size -= u->max_pruned_size;
1053         } else {
1054                 /* Reserve 5% memory in case the background free() are slower
1055                  * than the concurrent allocations. */
1056                 u->max_tree_size -= u->max_tree_size / 20;
1057         }
1058
1059         if (!u->prior)
1060                 u->prior = uct_prior_init(NULL, b, u);
1061
1062         if (!u->playout)
1063                 u->playout = playout_moggy_init(NULL, b, u->jdict);
1064         if (!u->playout->debug_level)
1065                 u->playout->debug_level = u->debug_level;
1066
1067         if (u->want_pat && !pat_setup)
1068                 patterns_init(&u->pat, NULL, false, true, spatial_dict_filename );
1069
1070         u->ownermap.map = malloc2(board_size2(b) * sizeof(u->ownermap.map[0]));
1071
1072         if (u->slave) {
1073                 if (!u->stats_hbits) u->stats_hbits = DEFAULT_STATS_HBITS;
1074                 if (!u->shared_nodes) u->shared_nodes = DEFAULT_SHARED_NODES;
1075                 assert(u->shared_levels * board_bits2(b) <= 8 * (int)sizeof(path_t));
1076         }
1077
1078         if (!u->dynkomi)
1079                 u->dynkomi = uct_dynkomi_init_linear(u, NULL, b);
1080
1081         /* Some things remain uninitialized for now - the opening tbook
1082          * is not loaded and the tree not set up. */
1083         /* This will be initialized in setup_state() at the first move
1084          * received/requested. This is because right now we are not aware
1085          * about any komi or handicap setup and such. */
1086
1087         return u;
1088 }
1089
1090 struct engine *
1091 engine_uct_init(char *arg, struct board *b)
1092 {
1093         struct uct *u = uct_state_init(arg, b);
1094         struct engine *e = calloc2(1, sizeof(struct engine));
1095         e->name = "UCT";
1096         e->printhook = uct_printhook_ownermap;
1097         e->notify_play = uct_notify_play;
1098         e->chat = uct_chat;
1099         e->undo = uct_undo;
1100         e->result = uct_result;
1101         e->genmove = uct_genmove;
1102         e->genmoves = uct_genmoves;
1103         e->evaluate = uct_evaluate;
1104         e->dead_group_list = uct_dead_group_list;
1105         e->done = uct_done;
1106         e->data = u;
1107         if (u->slave)
1108                 e->notify = uct_notify;
1109
1110         const char banner[] = "If you believe you have won but I am still playing, "
1111                 "please help me understand by capturing all dead stones. "
1112                 "Anyone can send me 'winrate' in private chat to get my assessment of the position.";
1113         if (!u->banner) u->banner = "";
1114         e->comment = malloc2(sizeof(banner) + strlen(u->banner) + 1);
1115         sprintf(e->comment, "%s %s", banner, u->banner);
1116
1117         return e;
1118 }