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);
  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(0))
  47                 fprintf(stderr, "Fresh board with random seed %lu\n", fast_getseed());
  48         //board_print(b, stderr);
  49         if (!u->no_tbook && b->moves == 0) {
  50                 assert(color == S_BLACK);
  51                 tree_load(u->t, b);
  52         }
  53 }
  54
  55 static void
  56 reset_state(struct uct *u)
  57 {
  58         assert(u->t);
  59         tree_done(u->t); u->t = NULL;
  60 }
  61
  62 static void
  63 setup_dynkomi(struct uct *u, struct board *b, enum stone to_play)
  64 {
  65         if (u->t->use_extra_komi && !u->pondering && u->dynkomi->permove)
  66                 u->t->extra_komi = u->dynkomi->permove(u->dynkomi, b, u->t);
  67         else if (!u->t->use_extra_komi)
  68                 u->t->extra_komi = 0;
  69 }
  70
  71 void
  72 uct_prepare_move(struct uct *u, struct board *b, enum stone color)
  73 {
  74         if (u->t) {
  75                 /* Verify that we have sane state. */
  76                 assert(b->es == u);
  77                 assert(u->t && b->moves);
  78                 if (color != stone_other(u->t->root_color)) {
  79                         fprintf(stderr, "Fatal: Non-alternating play detected %d %d\n",
  80                                 color, u->t->root_color);
  81                         exit(1);
  82                 }
  83                 uct_htable_reset(u->t);
  84
  85         } else {
  86                 /* We need fresh state. */
  87                 b->es = u;
  88                 setup_state(u, b, color);
  89         }
  90
  91         u->ownermap.playouts = 0;
  92         memset(u->ownermap.map, 0, board_size2(b) * sizeof(u->ownermap.map[0]));
  93         u->played_own = u->played_all = 0;
  94 }
  95
  96 static void
  97 dead_group_list(struct uct *u, struct board *b, struct move_queue *mq)
  98 {
  99         struct group_judgement gj;
 100         gj.thres = GJ_THRES;
 101         gj.gs = alloca(board_size2(b) * sizeof(gj.gs[0]));
 102         board_ownermap_judge_group(b, &u->ownermap, &gj);
 103         groups_of_status(b, &gj, GS_DEAD, mq);
 104 }
 105
 106 bool
 107 uct_pass_is_safe(struct uct *u, struct board *b, enum stone color, bool pass_all_alive)
 108 {
 109         if (u->ownermap.playouts < GJ_MINGAMES)
 110                 return false;
 111
 112         struct move_queue mq = { .moves = 0 };
 113         dead_group_list(u, b, &mq);
 114         if (pass_all_alive && mq.moves > 0)
 115                 return false; // We need to remove some dead groups first.
 116         return pass_is_safe(b, color, &mq);
 117 }
 118
 119 static char *
 120 uct_printhook_ownermap(struct board *board, coord_t c, char *s, char *end)
 121 {
 122         struct uct *u = board->es;
 123         if (!u) {
 124                 strcat(s, ". ");
 125                 return s + 2;
 126         }
 127         const char chr[] = ":XO,"; // dame, black, white, unclear
 128         const char chm[] = ":xo,";
 129         char ch = chr[board_ownermap_judge_point(&u->ownermap, c, GJ_THRES)];
 130         if (ch == ',') { // less precise estimate then?
 131                 ch = chm[board_ownermap_judge_point(&u->ownermap, c, 0.67)];
 132         }
 133         s += snprintf(s, end - s, "%c ", ch);
 134         return s;
 135 }
 136
 137 static char *
 138 uct_notify_play(struct engine *e, struct board *b, struct move *m)
 139 {
 140         struct uct *u = e->data;
 141         if (!u->t) {
 142                 /* No state, create one - this is probably game beginning
 143                  * and we need to load the opening tbook right now. */
 144                 uct_prepare_move(u, b, m->color);
 145                 assert(u->t);
 146         }
 147
 148         /* Stop pondering, required by tree_promote_at() */
 149         uct_pondering_stop(u);
 150         if (UDEBUGL(2) && u->slave)
 151                 tree_dump(u->t, u->dumpthres);
 152
 153         if (is_resign(m->coord)) {
 154                 /* Reset state. */
 155                 reset_state(u);
 156                 return NULL;
 157         }
 158
 159         /* Promote node of the appropriate move to the tree root. */
 160         assert(u->t->root);
 161         if (!tree_promote_at(u->t, b, m->coord)) {
 162                 if (UDEBUGL(0))
 163                         fprintf(stderr, "Warning: Cannot promote move node! Several play commands in row?\n");
 164                 reset_state(u);
 165                 return NULL;
 166         }
 167
 168         /* If we are a slave in a distributed engine, start pondering once
 169          * we know which move we actually played. See uct_genmove() about
 170          * the check for pass. */
 171         if (u->pondering_opt && u->slave && m->color == u->my_color && !is_pass(m->coord))
 172                 uct_pondering_start(u, b, u->t, stone_other(m->color));
 173
 174         return NULL;
 175 }
 176
 177 static char *
 178 uct_undo(struct engine *e, struct board *b)
 179 {
 180         struct uct *u = e->data;
 181
 182         if (!u->t) return NULL;
 183         uct_pondering_stop(u);
 184         reset_state(u);
 185         return NULL;
 186 }
 187
 188 static char *
 189 uct_result(struct engine *e, struct board *b)
 190 {
 191         struct uct *u = e->data;
 192         static char reply[1024];
 193
 194         if (!u->t)
 195                 return NULL;
 196         enum stone color = u->t->root_color;
 197         struct tree_node *n = u->t->root;
 198         snprintf(reply, 1024, "%s %s %d %.2f %.1f",
 199                  stone2str(color), coord2sstr(n->coord, b),
 200                  n->u.playouts, tree_node_get_value(u->t, -1, n->u.value),
 201                  u->t->use_extra_komi ? u->t->extra_komi : 0);
 202         return reply;
 203 }
 204
 205 static char *
 206 uct_chat(struct engine *e, struct board *b, char *cmd)
 207 {
 208         struct uct *u = e->data;
 209         static char reply[1024];
 210
 211         cmd += strspn(cmd, " \n\t");
 212         if (!strncasecmp(cmd, "winrate", 7)) {
 213                 if (!u->t)
 214                         return "no game context (yet?)";
 215                 enum stone color = u->t->root_color;
 216                 struct tree_node *n = u->t->root;
 217                 snprintf(reply, 1024, "In %d playouts at %d threads, %s %s can win with %.2f%% probability",
 218                          n->u.playouts, u->threads, stone2str(color), coord2sstr(n->coord, b),
 219                          tree_node_get_value(u->t, -1, n->u.value) * 100);
 220                 if (u->t->use_extra_komi && abs(u->t->extra_komi) >= 0.5) {
 221                         sprintf(reply + strlen(reply), ", while self-imposing extra komi %.1f",
 222                                 u->t->extra_komi);
 223                 }
 224                 strcat(reply, ".");
 225                 return reply;
 226         }
 227         return NULL;
 228 }
 229
 230 static void
 231 uct_dead_group_list(struct engine *e, struct board *b, struct move_queue *mq)
 232 {
 233         struct uct *u = e->data;
 234
 235         /* This means the game is probably over, no use pondering on. */
 236         uct_pondering_stop(u);
 237
 238         if (u->pass_all_alive)
 239                 return; // no dead groups
 240
 241         bool mock_state = false;
 242
 243         if (!u->t) {
 244                 /* No state, but we cannot just back out - we might
 245                  * have passed earlier, only assuming some stones are
 246                  * dead, and then re-connected, only to lose counting
 247                  * when all stones are assumed alive. */
 248                 uct_prepare_move(u, b, S_BLACK); assert(u->t);
 249                 mock_state = true;
 250         }
 251         /* Make sure the ownermap is well-seeded. */
 252         while (u->ownermap.playouts < GJ_MINGAMES)
 253                 uct_playout(u, b, S_BLACK, u->t);
 254         /* Show the ownermap: */
 255         if (DEBUGL(2))
 256                 board_print_custom(b, stderr, uct_printhook_ownermap);
 257
 258         dead_group_list(u, b, mq);
 259
 260         if (mock_state) {
 261                 /* Clean up the mock state in case we will receive
 262                  * a genmove; we could get a non-alternating-move
 263                  * error from uct_prepare_move() in that case otherwise. */
 264                 reset_state(u);
 265         }
 266 }
 267
 268 static void
 269 playout_policy_done(struct playout_policy *p)
 270 {
 271         if (p->done) p->done(p);
 272         if (p->data) free(p->data);
 273         free(p);
 274 }
 275
 276 static void
 277 uct_done(struct engine *e)
 278 {
 279         /* This is called on engine reset, especially when clear_board
 280          * is received and new game should begin. */
 281         struct uct *u = e->data;
 282         uct_pondering_stop(u);
 283         if (u->t) reset_state(u);
 284         free(u->ownermap.map);
 285
 286         free(u->policy);
 287         free(u->random_policy);
 288         playout_policy_done(u->playout);
 289         uct_prior_done(u->prior);
 290         joseki_done(u->jdict);
 291         pluginset_done(u->plugins);
 292 }
 293
 294
 295
 296 /* Run time-limited MCTS search on foreground. */
 297 static int
 298 uct_search(struct uct *u, struct board *b, struct time_info *ti, enum stone color, struct tree *t)
 299 {
 300         struct uct_search_state s;
 301         uct_search_start(u, b, color, t, ti, &s);
 302         if (UDEBUGL(2) && s.base_playouts > 0)
 303                 fprintf(stderr, "<pre-simulated %d games>\n", s.base_playouts);
 304
 305         /* The search tree is ctx->t. This is currently == . It is important
 306          * to reference ctx->t directly since the
 307          * thread manager will swap the tree pointer asynchronously. */
 308
 309         /* Now, just periodically poll the search tree. */
 310         /* Note that in case of TD_GAMES, threads will terminate independently
 311          * of the uct_search_check_stop() signalization. */
 312         while (1) {
 313                 time_sleep(TREE_BUSYWAIT_INTERVAL);
 314                 /* TREE_BUSYWAIT_INTERVAL should never be less than desired time, or the
 315                  * time control is broken. But if it happens to be less, we still search
 316                  * at least 100ms otherwise the move is completely random. */
 317
 318                 int i = uct_search_games(&s);
 319                 /* Print notifications etc. */
 320                 uct_search_progress(u, b, color, t, ti, &s, i);
 321                 /* Check if we should stop the search. */
 322                 if (uct_search_check_stop(u, b, color, t, ti, &s, i))
 323                         break;
 324         }
 325
 326         struct uct_thread_ctx *ctx = uct_search_stop();
 327         if (UDEBUGL(2)) tree_dump(t, u->dumpthres);
 328         if (UDEBUGL(2))
 329                 fprintf(stderr, "(avg score %f/%d value %f/%d)\n",
 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 = { t->root->coord, 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         /* Moreover, we do not use extra komi at the game end - we are not
 402          * to fool ourselves at this point. */
 403         if (board_estimated_moves_left(b) <= MIN_MOVES_LEFT)
 404                 u->t->use_extra_komi = false;
 405         setup_dynkomi(u, b, color);
 406
 407         if (b->rules == RULES_JAPANESE)
 408                 u->territory_scoring = true;
 409
 410         /* Make pessimistic assumption about komi for Japanese rules to
 411          * avoid losing by 0.5 when winning by 0.5 with Chinese rules.
 412          * The rules usually give the same winner if the integer part of komi
 413          * is odd so we adjust the komi only if it is even (for a board of
 414          * odd size). We are not trying  to get an exact evaluation for rare
 415          * cases of seki. For details see http://home.snafu.de/jasiek/parity.html */
 416         if (u->territory_scoring && (((int)floor(b->komi) + board_size(b)) & 1)) {
 417                 b->komi += (color == S_BLACK ? 1.0 : -1.0);
 418                 if (UDEBUGL(0))
 419                         fprintf(stderr, "Setting komi to %.1f assuming Japanese rules\n",
 420                                 b->komi);
 421         }
 422 }
 423
 424 static coord_t *
 425 uct_genmove(struct engine *e, struct board *b, struct time_info *ti, enum stone color, bool pass_all_alive)
 426 {
 427         double start_time = time_now();
 428         struct uct *u = e->data;
 429         uct_pondering_stop(u);
 430         uct_genmove_setup(u, b, color);
 431
 432         /* Start the Monte Carlo Tree Search! */
 433         int base_playouts = u->t->root->u.playouts;
 434         int played_games = uct_search(u, b, ti, color, u->t);
 435
 436         coord_t best_coord;
 437         struct tree_node *best;
 438         best = uct_search_result(u, b, color, pass_all_alive, played_games, base_playouts, &best_coord);
 439
 440         if (UDEBUGL(2)) {
 441                 double time = time_now() - start_time + 0.000001; /* avoid divide by zero */
 442                 fprintf(stderr, "genmove in %0.2fs (%d games/s, %d games/s/thread)\n",
 443                         time, (int)(played_games/time), (int)(played_games/time/u->threads));
 444         }
 445
 446         if (!best) {
 447                 /* Pass or resign. */
 448                 reset_state(u);
 449                 return coord_copy(best_coord);
 450         }
 451         tree_promote_node(u->t, &best);
 452
 453         /* After a pass, pondering is harmful for two reasons:
 454          * (i) We might keep pondering even when the game is over.
 455          * Of course this is the case for opponent resign as well.
 456          * (ii) More importantly, the ownermap will get skewed since
 457          * the UCT will start cutting off any playouts. */
 458         if (u->pondering_opt && !is_pass(best->coord)) {
 459                 uct_pondering_start(u, b, u->t, stone_other(color));
 460         }
 461         return coord_copy(best_coord);
 462 }
 463
 464
 465 bool
 466 uct_gentbook(struct engine *e, struct board *b, struct time_info *ti, enum stone color)
 467 {
 468         struct uct *u = e->data;
 469         if (!u->t) uct_prepare_move(u, b, color);
 470         assert(u->t);
 471
 472         if (ti->dim == TD_GAMES) {
 473                 /* Don't count in games that already went into the tbook. */
 474                 ti->len.games += u->t->root->u.playouts;
 475         }
 476         uct_search(u, b, ti, color, u->t);
 477
 478         assert(ti->dim == TD_GAMES);
 479         tree_save(u->t, b, ti->len.games / 100);
 480
 481         return true;
 482 }
 483
 484 void
 485 uct_dumptbook(struct engine *e, struct board *b, enum stone color)
 486 {
 487         struct uct *u = e->data;
 488         struct tree *t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
 489                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, 0);
 490         tree_load(t, b);
 491         tree_dump(t, 0);
 492         tree_done(t);
 493 }
 494
 495
 496 floating_t
 497 uct_evaluate(struct engine *e, struct board *b, struct time_info *ti, coord_t c, enum stone color)
 498 {
 499         struct uct *u = e->data;
 500
 501         struct board b2;
 502         board_copy(&b2, b);
 503         struct move m = { c, color };
 504         int res = board_play(&b2, &m);
 505         if (res < 0)
 506                 return NAN;
 507         color = stone_other(color);
 508
 509         if (u->t) reset_state(u);
 510         uct_prepare_move(u, &b2, color);
 511         assert(u->t);
 512
 513         floating_t bestval;
 514         uct_search(u, &b2, ti, color, u->t);
 515         struct tree_node *best = u->policy->choose(u->policy, u->t->root, &b2, color, resign);
 516         if (!best) {
 517                 bestval = NAN; // the opponent has no reply!
 518         } else {
 519                 bestval = tree_node_get_value(u->t, 1, best->u.value);
 520         }
 521
 522         reset_state(u); // clean our junk
 523
 524         return isnan(bestval) ? NAN : 1.0f - bestval;
 525 }
 526
 527
 528 struct uct *
 529 uct_state_init(char *arg, struct board *b)
 530 {
 531         struct uct *u = calloc2(1, sizeof(struct uct));
 532
 533         u->debug_level = debug_level;
 534         u->gamelen = MC_GAMELEN;
 535         u->resign_threshold = 0.2;
 536         u->sure_win_threshold = 0.85;
 537         u->mercymin = 0;
 538         u->significant_threshold = 50;
 539         u->expand_p = 2;
 540         u->dumpthres = 1000;
 541         u->playout_amaf = true;
 542         u->playout_amaf_nakade = false;
 543         u->amaf_prior = false;
 544         u->max_tree_size = 1408ULL * 1048576;
 545         u->fast_alloc = true;
 546         u->pruning_threshold = 0;
 547
 548         u->threads = 1;
 549         u->thread_model = TM_TREEVL;
 550         u->virtual_loss = 1;
 551
 552         u->fuseki_end = 20; // max time at 361*20% = 72 moves (our 36th move, still 99 to play)
 553         u->yose_start = 40; // (100-40-25)*361/100/2 = 63 moves still to play by us then
 554         u->bestr_ratio = 0.02;
 555         // 2.5 is clearly too much, but seems to compensate well for overly stern time allocations.
 556         // TODO: Further tuning and experiments with better time allocation schemes.
 557         u->best2_ratio = 2.5;
 558         u->max_maintime_ratio = 8.0;
 559
 560         u->val_scale = 0.04; u->val_points = 40;
 561         u->dynkomi_interval = 1000;
 562         u->dynkomi_mask = S_BLACK | S_WHITE;
 563
 564         u->tenuki_d = 4;
 565         u->local_tree_aging = 80;
 566         u->local_tree_allseq = 1;
 567         u->local_tree_rootseqval = 1;
 568         u->local_tree_depth_decay = 1.5;
 569
 570         u->plugins = pluginset_init(b);
 571
 572         u->jdict = joseki_load(b->size);
 573
 574         if (arg) {
 575                 char *optspec, *next = arg;
 576                 while (*next) {
 577                         optspec = next;
 578                         next += strcspn(next, ",");
 579                         if (*next) { *next++ = 0; } else { *next = 0; }
 580
 581                         char *optname = optspec;
 582                         char *optval = strchr(optspec, '=');
 583                         if (optval) *optval++ = 0;
 584
 585                         /** Basic options */
 586
 587                         if (!strcasecmp(optname, "debug")) {
 588                                 if (optval)
 589                                         u->debug_level = atoi(optval);
 590                                 else
 591                                         u->debug_level++;
 592                         } else if (!strcasecmp(optname, "reporting") && optval) {
 593                                 /* The format of output for detailed progress
 594                                  * information (such as current best move and
 595                                  * its value, etc.). */
 596                                 if (!strcasecmp(optval, "text")) {
 597                                         /* Plaintext traditional output. */
 598                                         u->reporting = UR_TEXT;
 599                                 } else {
 600                                         fprintf(stderr, "UCT: Invalid reporting format %s\n", optval);
 601                                         exit(1);
 602                                 }
 603                         } else if (!strcasecmp(optname, "dumpthres") && optval) {
 604                                 /* When dumping the UCT tree on output, include
 605                                  * nodes with at least this many playouts.
 606                                  * (This value is re-scaled "intelligently"
 607                                  * in case of very large trees.) */
 608                                 u->dumpthres = atoi(optval);
 609                         } else if (!strcasecmp(optname, "resign_threshold") && optval) {
 610                                 /* Resign when this ratio of games is lost
 611                                  * after GJ_MINGAMES sample is taken. */
 612                                 u->resign_threshold = atof(optval);
 613                         } else if (!strcasecmp(optname, "sure_win_threshold") && optval) {
 614                                 /* Stop reading when this ratio of games is won
 615                                  * after PLAYOUT_EARLY_BREAK_MIN sample is
 616                                  * taken. (Prevents stupid time losses,
 617                                  * friendly to human opponents.) */
 618                                 u->sure_win_threshold = atof(optval);
 619                         } else if (!strcasecmp(optname, "force_seed") && optval) {
 620                                 /* Set RNG seed at the tree setup. */
 621                                 u->force_seed = atoi(optval);
 622                         } else if (!strcasecmp(optname, "no_tbook")) {
 623                                 /* Disable UCT opening tbook. */
 624                                 u->no_tbook = true;
 625                         } else if (!strcasecmp(optname, "pass_all_alive")) {
 626                                 /* Whether to consider passing only after all
 627                                  * dead groups were removed from the board;
 628                                  * this is like all genmoves are in fact
 629                                  * kgs-genmove_cleanup. */
 630                                 u->pass_all_alive = !optval || atoi(optval);
 631                         } else if (!strcasecmp(optname, "territory_scoring")) {
 632                                 /* Use territory scoring (default is area scoring).
 633                                  * An explicit kgs-rules command overrides this. */
 634                                 u->territory_scoring = !optval || atoi(optval);
 635                         } else if (!strcasecmp(optname, "banner") && optval) {
 636                                 /* Additional banner string. This must come as the
 637                                  * last engine parameter. */
 638                                 if (*next) *--next = ',';
 639                                 u->banner = strdup(optval);
 640                                 break;
 641                         } else if (!strcasecmp(optname, "plugin") && optval) {
 642                                 /* Load an external plugin; filename goes before the colon,
 643                                  * extra arguments after the colon. */
 644                                 char *pluginarg = strchr(optval, ':');
 645                                 if (pluginarg)
 646                                         *pluginarg++ = 0;
 647                                 plugin_load(u->plugins, optval, pluginarg);
 648
 649                         /** UCT behavior and policies */
 650
 651                         } else if ((!strcasecmp(optname, "policy")
 652                                 /* Node selection policy. ucb1amaf is the
 653                                  * default policy implementing RAVE, while
 654                                  * ucb1 is the simple exploration/exploitation
 655                                  * policy. Policies can take further extra
 656                                  * options. */
 657                                     || !strcasecmp(optname, "random_policy")) && optval) {
 658                                 /* A policy to be used randomly with small
 659                                  * chance instead of the default policy. */
 660                                 char *policyarg = strchr(optval, ':');
 661                                 struct uct_policy **p = !strcasecmp(optname, "policy") ? &u->policy : &u->random_policy;
 662                                 if (policyarg)
 663                                         *policyarg++ = 0;
 664                                 if (!strcasecmp(optval, "ucb1")) {
 665                                         *p = policy_ucb1_init(u, policyarg);
 666                                 } else if (!strcasecmp(optval, "ucb1amaf")) {
 667                                         *p = policy_ucb1amaf_init(u, policyarg);
 668                                 } else {
 669                                         fprintf(stderr, "UCT: Invalid tree policy %s\n", optval);
 670                                         exit(1);
 671                                 }
 672                         } else if (!strcasecmp(optname, "playout") && optval) {
 673                                 /* Random simulation (playout) policy.
 674                                  * moggy is the default policy with large
 675                                  * amount of domain-specific knowledge and
 676                                  * heuristics. light is a simple uniformly
 677                                  * random move selection policy. */
 678                                 char *playoutarg = strchr(optval, ':');
 679                                 if (playoutarg)
 680                                         *playoutarg++ = 0;
 681                                 if (!strcasecmp(optval, "moggy")) {
 682                                         u->playout = playout_moggy_init(playoutarg, b, u->jdict);
 683                                 } else if (!strcasecmp(optval, "light")) {
 684                                         u->playout = playout_light_init(playoutarg, b);
 685                                 } else {
 686                                         fprintf(stderr, "UCT: Invalid playout policy %s\n", optval);
 687                                         exit(1);
 688                                 }
 689                         } else if (!strcasecmp(optname, "prior") && optval) {
 690                                 /* Node priors policy. When expanding a node,
 691                                  * it will seed node values heuristically
 692                                  * (most importantly, based on playout policy
 693                                  * opinion, but also with regard to other
 694                                  * things). See uct/prior.c for details.
 695                                  * Use prior=eqex=0 to disable priors. */
 696                                 u->prior = uct_prior_init(optval, b);
 697                         } else if (!strcasecmp(optname, "mercy") && optval) {
 698                                 /* Minimal difference of black/white captures
 699                                  * to stop playout - "Mercy Rule". Speeds up
 700                                  * hopeless playouts at the expense of some
 701                                  * accuracy. */
 702                                 u->mercymin = atoi(optval);
 703                         } else if (!strcasecmp(optname, "gamelen") && optval) {
 704                                 /* Maximum length of single simulation
 705                                  * in moves. */
 706                                 u->gamelen = atoi(optval);
 707                         } else if (!strcasecmp(optname, "expand_p") && optval) {
 708                                 /* Expand UCT nodes after it has been
 709                                  * visited this many times. */
 710                                 u->expand_p = atoi(optval);
 711                         } else if (!strcasecmp(optname, "random_policy_chance") && optval) {
 712                                 /* If specified (N), with probability 1/N, random_policy policy
 713                                  * descend is used instead of main policy descend; useful
 714                                  * if specified policy (e.g. UCB1AMAF) can make unduly biased
 715                                  * choices sometimes, you can fall back to e.g.
 716                                  * random_policy=UCB1. */
 717                                 u->random_policy_chance = atoi(optval);
 718
 719                         /** General AMAF behavior */
 720                         /* (Only relevant if the policy supports AMAF.
 721                          * More variables can be tuned as policy
 722                          * parameters.) */
 723
 724                         } else if (!strcasecmp(optname, "playout_amaf")) {
 725                                 /* Whether to include random playout moves in
 726                                  * AMAF as well. (Otherwise, only tree moves
 727                                  * are included in AMAF. Of course makes sense
 728                                  * only in connection with an AMAF policy.) */
 729                                 /* with-without: 55.5% (+-4.1) */
 730                                 if (optval && *optval == '0')
 731                                         u->playout_amaf = false;
 732                                 else
 733                                         u->playout_amaf = true;
 734                         } else if (!strcasecmp(optname, "playout_amaf_nakade")) {
 735                                 /* Whether to include nakade moves from playouts
 736                                  * in the AMAF statistics; this tends to nullify
 737                                  * the playout_amaf effect by adding too much
 738                                  * noise. */
 739                                 if (optval && *optval == '0')
 740                                         u->playout_amaf_nakade = false;
 741                                 else
 742                                         u->playout_amaf_nakade = true;
 743                         } else if (!strcasecmp(optname, "playout_amaf_cutoff") && optval) {
 744                                 /* Keep only first N% of playout stage AMAF
 745                                  * information. */
 746                                 u->playout_amaf_cutoff = atoi(optval);
 747                         } else if (!strcasecmp(optname, "amaf_prior") && optval) {
 748                                 /* In node policy, consider prior values
 749                                  * part of the real result term or part
 750                                  * of the AMAF term? */
 751                                 u->amaf_prior = atoi(optval);
 752
 753                         /** Performance and memory management */
 754
 755                         } else if (!strcasecmp(optname, "threads") && optval) {
 756                                 /* By default, Pachi will run with only single
 757                                  * tree search thread! */
 758                                 u->threads = atoi(optval);
 759                         } else if (!strcasecmp(optname, "thread_model") && optval) {
 760                                 if (!strcasecmp(optval, "tree")) {
 761                                         /* Tree parallelization - all threads
 762                                          * grind on the same tree. */
 763                                         u->thread_model = TM_TREE;
 764                                         u->virtual_loss = 0;
 765                                 } else if (!strcasecmp(optval, "treevl")) {
 766                                         /* Tree parallelization, but also
 767                                          * with virtual losses - this discou-
 768                                          * rages most threads choosing the
 769                                          * same tree branches to read. */
 770                                         u->thread_model = TM_TREEVL;
 771                                 } else {
 772                                         fprintf(stderr, "UCT: Invalid thread model %s\n", optval);
 773                                         exit(1);
 774                                 }
 775                         } else if (!strcasecmp(optname, "virtual_loss")) {
 776                                 /* Number of virtual losses added before evaluating a node. */
 777                                 u->virtual_loss = !optval || atoi(optval);
 778                         } else if (!strcasecmp(optname, "pondering")) {
 779                                 /* Keep searching even during opponent's turn. */
 780                                 u->pondering_opt = !optval || atoi(optval);
 781                         } else if (!strcasecmp(optname, "max_tree_size") && optval) {
 782                                 /* Maximum amount of memory [MiB] consumed by the move tree.
 783                                  * For fast_alloc it includes the temp tree used for pruning.
 784                                  * Default is 3072 (3 GiB). */
 785                                 u->max_tree_size = atol(optval) * 1048576;
 786                         } else if (!strcasecmp(optname, "fast_alloc")) {
 787                                 u->fast_alloc = !optval || atoi(optval);
 788                         } else if (!strcasecmp(optname, "pruning_threshold") && optval) {
 789                                 /* Force pruning at beginning of a move if the tree consumes
 790                                  * more than this [MiB]. Default is 10% of max_tree_size.
 791                                  * Increase to reduce pruning time overhead if memory is plentiful.
 792                                  * This option is meaningful only for fast_alloc. */
 793                                 u->pruning_threshold = atol(optval) * 1048576;
 794
 795                         /** Time control */
 796
 797                         } else if (!strcasecmp(optname, "best2_ratio") && optval) {
 798                                 /* If set, prolong simulating while
 799                                  * first_best/second_best playouts ratio
 800                                  * is less than best2_ratio. */
 801                                 u->best2_ratio = atof(optval);
 802                         } else if (!strcasecmp(optname, "bestr_ratio") && optval) {
 803                                 /* If set, prolong simulating while
 804                                  * best,best_best_child values delta
 805                                  * is more than bestr_ratio. */
 806                                 u->bestr_ratio = atof(optval);
 807                         } else if (!strcasecmp(optname, "max_maintime_ratio") && optval) {
 808                                 /* If set and while not in byoyomi, prolong simulating no more than
 809                                  * max_maintime_ratio times the normal desired thinking time. */
 810                                 u->max_maintime_ratio = atof(optval);
 811                         } else if (!strcasecmp(optname, "fuseki_end") && optval) {
 812                                 /* At the very beginning it's not worth thinking
 813                                  * too long because the playout evaluations are
 814                                  * very noisy. So gradually increase the thinking
 815                                  * time up to maximum when fuseki_end percent
 816                                  * of the board has been played.
 817                                  * This only applies if we are not in byoyomi. */
 818                                 u->fuseki_end = atoi(optval);
 819                         } else if (!strcasecmp(optname, "yose_start") && optval) {
 820                                 /* When yose_start percent of the board has been
 821                                  * played, or if we are in byoyomi, stop spending
 822                                  * more time and spread the remaining time
 823                                  * uniformly.
 824                                  * Between fuseki_end and yose_start, we spend
 825                                  * a constant proportion of the remaining time
 826                                  * on each move. (yose_start should actually
 827                                  * be much earlier than when real yose start,
 828                                  * but "yose" is a good short name to convey
 829                                  * the idea.) */
 830                                 u->yose_start = atoi(optval);
 831
 832                         /** Dynamic komi */
 833
 834                         } else if (!strcasecmp(optname, "dynkomi") && optval) {
 835                                 /* Dynamic komi approach; there are multiple
 836                                  * ways to adjust komi dynamically throughout
 837                                  * play. We currently support two: */
 838                                 char *dynkomiarg = strchr(optval, ':');
 839                                 if (dynkomiarg)
 840                                         *dynkomiarg++ = 0;
 841                                 if (!strcasecmp(optval, "none")) {
 842                                         u->dynkomi = uct_dynkomi_init_none(u, dynkomiarg, b);
 843                                 } else if (!strcasecmp(optval, "linear")) {
 844                                         /* You should set dynkomi_mask=1
 845                                          * since this doesn't work well
 846                                          * for white handicaps! */
 847                                         u->dynkomi = uct_dynkomi_init_linear(u, dynkomiarg, b);
 848                                 } else if (!strcasecmp(optval, "adaptive")) {
 849                                         /* There are many more knobs to
 850                                          * crank - see uct/dynkomi.c. */
 851                                         u->dynkomi = uct_dynkomi_init_adaptive(u, dynkomiarg, b);
 852                                 } else {
 853                                         fprintf(stderr, "UCT: Invalid dynkomi mode %s\n", optval);
 854                                         exit(1);
 855                                 }
 856                         } else if (!strcasecmp(optname, "dynkomi_mask") && optval) {
 857                                 /* Bitmask of colors the player must be
 858                                  * for dynkomi be applied; you may want
 859                                  * to use dynkomi_mask=3 to allow dynkomi
 860                                  * even in games where Pachi is white. */
 861                                 u->dynkomi_mask = atoi(optval);
 862                         } else if (!strcasecmp(optname, "dynkomi_interval") && optval) {
 863                                 /* If non-zero, re-adjust dynamic komi
 864                                  * throughout a single genmove reading,
 865                                  * roughly every N simulations. */
 866                                 /* XXX: Does not work with tree
 867                                  * parallelization. */
 868                                 u->dynkomi_interval = atoi(optval);
 869
 870                         /** Node value result scaling */
 871
 872                         } else if (!strcasecmp(optname, "val_scale") && optval) {
 873                                 /* How much of the game result value should be
 874                                  * influenced by win size. Zero means it isn't. */
 875                                 u->val_scale = atof(optval);
 876                         } else if (!strcasecmp(optname, "val_points") && optval) {
 877                                 /* Maximum size of win to be scaled into game
 878                                  * result value. Zero means boardsize^2. */
 879                                 u->val_points = atoi(optval) * 2; // result values are doubled
 880                         } else if (!strcasecmp(optname, "val_extra")) {
 881                                 /* If false, the score coefficient will be simply
 882                                  * added to the value, instead of scaling the result
 883                                  * coefficient because of it. */
 884                                 u->val_extra = !optval || atoi(optval);
 885
 886                         /** Local trees */
 887                         /* (Purely experimental. Does not work - yet!) */
 888
 889                         } else if (!strcasecmp(optname, "local_tree") && optval) {
 890                                 /* Whether to bias exploration by local tree values
 891                                  * (must be supported by the used policy).
 892                                  * 0: Don't.
 893                                  * 1: Do, value = result.
 894                                  * Try to temper the result:
 895                                  * 2: Do, value = 0.5+(result-expected)/2.
 896                                  * 3: Do, value = 0.5+bzz((result-expected)^2).
 897                                  * 4: Do, value = 0.5+sqrt(result-expected)/2. */
 898                                 u->local_tree = atoi(optval);
 899                         } else if (!strcasecmp(optname, "tenuki_d") && optval) {
 900                                 /* Tenuki distance at which to break the local tree. */
 901                                 u->tenuki_d = atoi(optval);
 902                                 if (u->tenuki_d > TREE_NODE_D_MAX + 1) {
 903                                         fprintf(stderr, "uct: tenuki_d must not be larger than TREE_NODE_D_MAX+1 %d\n", TREE_NODE_D_MAX + 1);
 904                                         exit(1);
 905                                 }
 906                         } else if (!strcasecmp(optname, "local_tree_aging") && optval) {
 907                                 /* How much to reduce local tree values between moves. */
 908                                 u->local_tree_aging = atof(optval);
 909                         } else if (!strcasecmp(optname, "local_tree_depth_decay") && optval) {
 910                                 /* With value x>0, during the descent the node
 911                                  * contributes 1/x^depth playouts in
 912                                  * the local tree. I.e., with x>1, nodes more
 913                                  * distant from local situation contribute more
 914                                  * than nodes near the root. */
 915                                 u->local_tree_depth_decay = atof(optval);
 916                         } else if (!strcasecmp(optname, "local_tree_allseq")) {
 917                                 /* If disabled, only complete sequences are stored
 918                                  * in the local tree. If this is on, also
 919                                  * subsequences starting at each move are stored. */
 920                                 u->local_tree_allseq = !optval || atoi(optval);
 921                         } else if (!strcasecmp(optname, "local_tree_rootseqval")) {
 922                                 /* If disabled, expected node value is computed by
 923                                  * summing up values through the whole descent.
 924                                  * If enabled, expected node value for
 925                                  * each sequence is the value at the root of the
 926                                  * sequence. */
 927                                 u->local_tree_rootseqval = !optval || atoi(optval);
 928
 929                         /** Other heuristics */
 930                         } else if (!strcasecmp(optname, "significant_threshold") && optval) {
 931                                 /* Some heuristics (XXX: none in mainline) rely
 932                                  * on the knowledge of the last "significant"
 933                                  * node in the descent. Such a node is
 934                                  * considered reasonably trustworthy to carry
 935                                  * some meaningful information in the values
 936                                  * of the node and its children. */
 937                                 u->significant_threshold = atoi(optval);
 938
 939                         /** Distributed engine slaves setup */
 940
 941                         } else if (!strcasecmp(optname, "slave")) {
 942                                 /* Act as slave for the distributed engine. */
 943                                 u->slave = !optval || atoi(optval);
 944                         } else if (!strcasecmp(optname, "shared_nodes") && optval) {
 945                                 /* Share at most shared_nodes between master and slave at each genmoves.
 946                                  * Must use the same value in master and slaves. */
 947                                 u->shared_nodes = atoi(optval);
 948                         } else if (!strcasecmp(optname, "shared_levels") && optval) {
 949                                 /* Share only nodes of level <= shared_levels. */
 950                                 u->shared_levels = atoi(optval);
 951                         } else if (!strcasecmp(optname, "stats_hbits") && optval) {
 952                                 /* Set hash table size to 2^stats_hbits for the shared stats. */
 953                                 u->stats_hbits = atoi(optval);
 954
 955                         } else {
 956                                 fprintf(stderr, "uct: Invalid engine argument %s or missing value\n", optname);
 957                                 exit(1);
 958                         }
 959                 }
 960         }
 961
 962         if (!u->policy)
 963                 u->policy = policy_ucb1amaf_init(u, NULL);
 964
 965         if (!!u->random_policy_chance ^ !!u->random_policy) {
 966                 fprintf(stderr, "uct: Only one of random_policy and random_policy_chance is set\n");
 967                 exit(1);
 968         }
 969
 970         if (!u->local_tree) {
 971                 /* No ltree aging. */
 972                 u->local_tree_aging = 1.0f;
 973         }
 974
 975         if (u->fast_alloc) {
 976                 if (u->pruning_threshold < u->max_tree_size / 10)
 977                         u->pruning_threshold = u->max_tree_size / 10;
 978                 if (u->pruning_threshold > u->max_tree_size / 2)
 979                         u->pruning_threshold = u->max_tree_size / 2;
 980
 981                 /* Limit pruning temp space to 20% of memory. Beyond this we discard
 982                  * the nodes and recompute them at the next move if necessary. */
 983                 u->max_pruned_size = u->max_tree_size / 5;
 984                 u->max_tree_size -= u->max_pruned_size;
 985         } else {
 986                 /* Reserve 5% memory in case the background free() are slower
 987                  * than the concurrent allocations. */
 988                 u->max_tree_size -= u->max_tree_size / 20;
 989         }
 990
 991         if (!u->prior)
 992                 u->prior = uct_prior_init(NULL, b);
 993
 994         if (!u->playout)
 995                 u->playout = playout_moggy_init(NULL, b, u->jdict);
 996         if (!u->playout->debug_level)
 997                 u->playout->debug_level = u->debug_level;
 998
 999         u->ownermap.map = malloc2(board_size2(b) * sizeof(u->ownermap.map[0]));
1000
1001         if (u->slave) {
1002                 if (!u->stats_hbits) u->stats_hbits = DEFAULT_STATS_HBITS;
1003                 if (!u->shared_nodes) u->shared_nodes = DEFAULT_SHARED_NODES;
1004                 assert(u->shared_levels * board_bits2(b) <= 8 * (int)sizeof(path_t));
1005         }
1006
1007         if (!u->dynkomi)
1008                 u->dynkomi = uct_dynkomi_init_adaptive(u, NULL, b);
1009
1010         /* Some things remain uninitialized for now - the opening tbook
1011          * is not loaded and the tree not set up. */
1012         /* This will be initialized in setup_state() at the first move
1013          * received/requested. This is because right now we are not aware
1014          * about any komi or handicap setup and such. */
1015
1016         return u;
1017 }
1018
1019 struct engine *
1020 engine_uct_init(char *arg, struct board *b)
1021 {
1022         struct uct *u = uct_state_init(arg, b);
1023         struct engine *e = calloc2(1, sizeof(struct engine));
1024         e->name = "UCT Engine";
1025         e->printhook = uct_printhook_ownermap;
1026         e->notify_play = uct_notify_play;
1027         e->chat = uct_chat;
1028         e->undo = uct_undo;
1029         e->result = uct_result;
1030         e->genmove = uct_genmove;
1031         e->genmoves = uct_genmoves;
1032         e->dead_group_list = uct_dead_group_list;
1033         e->done = uct_done;
1034         e->data = u;
1035         if (u->slave)
1036                 e->notify = uct_notify;
1037
1038         const char banner[] = "I'm playing UCT. When I'm losing, I will resign, "
1039                 "if I think I win, I play until you pass. "
1040                 "Anyone can send me 'winrate' in private chat to get my assessment of the position.";
1041         if (!u->banner) u->banner = "";
1042         e->comment = malloc2(sizeof(banner) + strlen(u->banner) + 1);
1043         sprintf(e->comment, "%s %s", banner, u->banner);
1044
1045         return e;
1046 }