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         u->t->extra_komi = u->saved_extra_komi;
  45         if (u->force_seed)
  46                 fast_srandom(u->force_seed);
  47         if (UDEBUGL(0))
  48                 fprintf(stderr, "Fresh board with random seed %lu\n", fast_getseed());
  49         //board_print(b, stderr);
  50         if (!u->no_tbook && b->moves == 0) {
  51                 assert(color == S_BLACK);
  52                 tree_load(u->t, b);
  53         }
  54 }
  55
  56 static void
  57 remove_tree(struct uct *u)
  58 {
  59         assert(u->t);
  60         tree_done(u->t); u->t = NULL;
  61 }
  62
  63 static void
  64 reset_state(struct uct *u)
  65 {
  66         u->saved_extra_komi = 0.0;
  67         remove_tree(u);
  68 }
  69
  70 static void
  71 reset_state_keep_komi(struct uct *u)
  72 {
  73         u->saved_extra_komi = u->t->extra_komi;
  74         remove_tree(u);
  75 }
  76
  77 static void
  78 setup_dynkomi(struct uct *u, struct board *b, enum stone to_play)
  79 {
  80         if (u->t->use_extra_komi && !u->pondering && u->dynkomi->permove)
  81                 u->t->extra_komi = u->dynkomi->permove(u->dynkomi, b, u->t);
  82         else if (!u->t->use_extra_komi)
  83                 u->t->extra_komi = 0;
  84 }
  85
  86 void
  87 uct_prepare_move(struct uct *u, struct board *b, enum stone color)
  88 {
  89         if (u->t) {
  90                 /* Verify that we have sane state. */
  91                 assert(b->es == u);
  92                 assert(u->t && b->moves);
  93                 if (color != stone_other(u->t->root_color)) {
  94                         fprintf(stderr, "Fatal: Non-alternating play detected %d %d\n",
  95                                 color, u->t->root_color);
  96                         exit(1);
  97                 }
  98                 uct_htable_reset(u->t);
  99
 100         } else {
 101                 /* We need fresh state. */
 102                 b->es = u;
 103                 setup_state(u, b, color);
 104         }
 105
 106         u->ownermap.playouts = 0;
 107         memset(u->ownermap.map, 0, board_size2(b) * sizeof(u->ownermap.map[0]));
 108         u->played_own = u->played_all = 0;
 109 }
 110
 111 static void
 112 dead_group_list(struct uct *u, struct board *b, struct move_queue *mq)
 113 {
 114         struct group_judgement gj;
 115         gj.thres = GJ_THRES;
 116         gj.gs = alloca(board_size2(b) * sizeof(gj.gs[0]));
 117         board_ownermap_judge_group(b, &u->ownermap, &gj);
 118         groups_of_status(b, &gj, GS_DEAD, mq);
 119 }
 120
 121 bool
 122 uct_pass_is_safe(struct uct *u, struct board *b, enum stone color, bool pass_all_alive)
 123 {
 124         if (u->ownermap.playouts < GJ_MINGAMES)
 125                 return false;
 126
 127         struct move_queue mq = { .moves = 0 };
 128         dead_group_list(u, b, &mq);
 129         if (pass_all_alive && mq.moves > 0)
 130                 return false; // We need to remove some dead groups first.
 131         return pass_is_safe(b, color, &mq);
 132 }
 133
 134 static char *
 135 uct_printhook_ownermap(struct board *board, coord_t c, char *s, char *end)
 136 {
 137         struct uct *u = board->es;
 138         if (!u) {
 139                 strcat(s, ". ");
 140                 return s + 2;
 141         }
 142         const char chr[] = ":XO,"; // dame, black, white, unclear
 143         const char chm[] = ":xo,";
 144         char ch = chr[board_ownermap_judge_point(&u->ownermap, c, GJ_THRES)];
 145         if (ch == ',') { // less precise estimate then?
 146                 ch = chm[board_ownermap_judge_point(&u->ownermap, c, 0.67)];
 147         }
 148         s += snprintf(s, end - s, "%c ", ch);
 149         return s;
 150 }
 151
 152 static char *
 153 uct_notify_play(struct engine *e, struct board *b, struct move *m)
 154 {
 155         struct uct *u = e->data;
 156         if (!u->t) {
 157                 /* No state, create one - this is probably game beginning
 158                  * and we need to load the opening tbook right now. */
 159                 uct_prepare_move(u, b, m->color);
 160                 assert(u->t);
 161         }
 162
 163         /* Stop pondering, required by tree_promote_at() */
 164         uct_pondering_stop(u);
 165         if (UDEBUGL(2) && u->slave) /* remove the u->slave test to show the pruned tree */
 166                 tree_dump(u->t, u->dumpthres, u->unlimited_tree_dump);
 167
 168         if (is_resign(m->coord)) {
 169                 /* Reset state. */
 170                 reset_state(u);
 171                 return NULL;
 172         }
 173
 174         /* Promote node of the appropriate move to the tree root. */
 175         assert(u->t->root);
 176         if (!tree_promote_at(u->t, b, m->coord)) {
 177                 /* The opponent move does not exist in the tree.  Maybe the playout count of our
 178                  * previous move node is so small that it did not get expanded to yield the opponent
 179                  * moves, or more likely, the pruning process done at the end of our previous move
 180                  * removed the opponent node; indeed, since the expansion process in pruning is
 181                  * depth-first, and any children failing allocation will cause the parent not to be
 182                  * expanded, upon a temp-tree overflow we often have only a root node left. */
 183                 if (UDEBUGL(0))
 184                         fprintf(stderr, "Warning: Cannot promote move node!\n");
 185                 reset_state_keep_komi(u);
 186                 return NULL;
 187         }
 188
 189         /* If we are a slave in a distributed engine, start pondering once
 190          * we know which move we actually played. See uct_genmove() about
 191          * the check for pass. */
 192         if (u->pondering_opt && u->slave && m->color == u->my_color && !is_pass(m->coord))
 193                 uct_pondering_start(u, b, u->t, stone_other(m->color));
 194
 195         return NULL;
 196 }
 197
 198 static char *
 199 uct_undo(struct engine *e, struct board *b)
 200 {
 201         struct uct *u = e->data;
 202
 203         if (!u->t) return NULL;
 204         uct_pondering_stop(u);
 205         reset_state(u);
 206         return NULL;
 207 }
 208
 209 static char *
 210 uct_result(struct engine *e, struct board *b)
 211 {
 212         struct uct *u = e->data;
 213         static char reply[1024];
 214
 215         if (!u->t)
 216                 return NULL;
 217         enum stone color = u->t->root_color;
 218         struct tree_node *n = u->t->root;
 219         snprintf(reply, 1024, "%s %s %d %.2f %.1f",
 220                  stone2str(color), coord2sstr(n->coord, b),
 221                  n->u.playouts, tree_node_get_value(u->t, -1, n->u.value),
 222                  u->t->use_extra_komi ? u->t->extra_komi : 0);
 223         return reply;
 224 }
 225
 226 static char *
 227 uct_chat(struct engine *e, struct board *b, char *cmd)
 228 {
 229         struct uct *u = e->data;
 230         static char reply[1024];
 231
 232         cmd += strspn(cmd, " \n\t");
 233         if (!strncasecmp(cmd, "winrate", 7)) {
 234                 if (!u->t)
 235                         return "no game context (yet?)";
 236                 enum stone color = u->t->root_color;
 237                 struct tree_node *n = u->t->root;
 238                 snprintf(reply, 1024, "In %d playouts at %d threads, %s %s can win with %.2f%% probability",
 239                          n->u.playouts, u->threads, stone2str(color), coord2sstr(n->coord, b),
 240                          tree_node_get_value(u->t, -1, n->u.value) * 100);
 241                 if (u->t->use_extra_komi && abs(u->t->extra_komi) >= 0.5) {
 242                         sprintf(reply + strlen(reply), ", while self-imposing extra komi %.1f",
 243                                 u->t->extra_komi);
 244                 }
 245                 strcat(reply, ".");
 246                 return reply;
 247         }
 248         return NULL;
 249 }
 250
 251 static void
 252 uct_dead_group_list(struct engine *e, struct board *b, struct move_queue *mq)
 253 {
 254         struct uct *u = e->data;
 255
 256         /* This means the game is probably over, no use pondering on. */
 257         uct_pondering_stop(u);
 258
 259         if (u->pass_all_alive)
 260                 return; // no dead groups
 261
 262         bool mock_state = false;
 263
 264         if (!u->t) {
 265                 /* No state, but we cannot just back out - we might
 266                  * have passed earlier, only assuming some stones are
 267                  * dead, and then re-connected, only to lose counting
 268                  * when all stones are assumed alive. */
 269                 uct_prepare_move(u, b, S_BLACK); assert(u->t);
 270                 mock_state = true;
 271         }
 272         /* Make sure the ownermap is well-seeded. */
 273         while (u->ownermap.playouts < GJ_MINGAMES)
 274                 uct_playout(u, b, S_BLACK, u->t);
 275         /* Show the ownermap: */
 276         if (DEBUGL(2))
 277                 board_print_custom(b, stderr, uct_printhook_ownermap);
 278
 279         dead_group_list(u, b, mq);
 280
 281         if (mock_state) {
 282                 /* Clean up the mock state in case we will receive
 283                  * a genmove; we could get a non-alternating-move
 284                  * error from uct_prepare_move() in that case otherwise. */
 285                 reset_state(u);
 286         }
 287 }
 288
 289 static void
 290 playout_policy_done(struct playout_policy *p)
 291 {
 292         if (p->done) p->done(p);
 293         if (p->data) free(p->data);
 294         free(p);
 295 }
 296
 297 static void
 298 uct_done(struct engine *e)
 299 {
 300         /* This is called on engine reset, especially when clear_board
 301          * is received and new game should begin. */
 302         struct uct *u = e->data;
 303         uct_pondering_stop(u);
 304         if (u->t) reset_state(u);
 305         free(u->ownermap.map);
 306
 307         free(u->policy);
 308         free(u->random_policy);
 309         playout_policy_done(u->playout);
 310         uct_prior_done(u->prior);
 311         joseki_done(u->jdict);
 312         pluginset_done(u->plugins);
 313 }
 314
 315
 316
 317 /* Run time-limited MCTS search on foreground. */
 318 static int
 319 uct_search(struct uct *u, struct board *b, struct time_info *ti, enum stone color, struct tree *t)
 320 {
 321         struct uct_search_state s;
 322         uct_search_start(u, b, color, t, ti, &s);
 323         if (UDEBUGL(2) && s.base_playouts > 0)
 324                 fprintf(stderr, "<pre-simulated %d games>\n", s.base_playouts);
 325
 326         /* The search tree is ctx->t. This is currently == . It is important
 327          * to reference ctx->t directly since the
 328          * thread manager will swap the tree pointer asynchronously. */
 329
 330         /* Now, just periodically poll the search tree. */
 331         /* Note that in case of TD_GAMES, threads will terminate independently
 332          * of the uct_search_check_stop() signalization. */
 333         while (1) {
 334                 time_sleep(TREE_BUSYWAIT_INTERVAL);
 335                 /* TREE_BUSYWAIT_INTERVAL should never be less than desired time, or the
 336                  * time control is broken. But if it happens to be less, we still search
 337                  * at least 100ms otherwise the move is completely random. */
 338
 339                 int i = uct_search_games(&s);
 340                 /* Print notifications etc. */
 341                 uct_search_progress(u, b, color, t, ti, &s, i);
 342                 /* Check if we should stop the search. */
 343                 if (uct_search_check_stop(u, b, color, t, ti, &s, i))
 344                         break;
 345         }
 346
 347         struct uct_thread_ctx *ctx = uct_search_stop();
 348         if (UDEBUGL(2)) tree_dump(t, u->dumpthres, u->unlimited_tree_dump);
 349         if (UDEBUGL(2))
 350                 fprintf(stderr, "(avg score %f/%d value %f/%d)\n",
 351                         u->dynkomi->score.value, u->dynkomi->score.playouts,
 352                         u->dynkomi->value.value, u->dynkomi->value.playouts);
 353         if (UDEBUGL(0))
 354                 uct_progress_status(u, t, color, ctx->games);
 355
 356         u->played_own += ctx->games;
 357         return ctx->games;
 358 }
 359
 360 /* Start pondering background with @color to play. */
 361 static void
 362 uct_pondering_start(struct uct *u, struct board *b0, struct tree *t, enum stone color)
 363 {
 364         if (UDEBUGL(1))
 365                 fprintf(stderr, "Starting to ponder with color %s\n", stone2str(stone_other(color)));
 366         u->pondering = true;
 367
 368         /* We need a local board copy to ponder upon. */
 369         struct board *b = malloc2(sizeof(*b)); board_copy(b, b0);
 370
 371         /* *b0 did not have the genmove'd move played yet. */
 372         struct move m = { t->root->coord, t->root_color };
 373         int res = board_play(b, &m);
 374         assert(res >= 0);
 375         setup_dynkomi(u, b, stone_other(m.color));
 376
 377         /* Start MCTS manager thread "headless". */
 378         static struct uct_search_state s;
 379         uct_search_start(u, b, color, t, NULL, &s);
 380 }
 381
 382 /* uct_search_stop() frontend for the pondering (non-genmove) mode, and
 383  * to stop the background search for a slave in the distributed engine. */
 384 void
 385 uct_pondering_stop(struct uct *u)
 386 {
 387         if (!thread_manager_running)
 388                 return;
 389
 390         /* Stop the thread manager. */
 391         struct uct_thread_ctx *ctx = uct_search_stop();
 392         if (UDEBUGL(1)) {
 393                 if (u->pondering) fprintf(stderr, "(pondering) ");
 394                 uct_progress_status(u, ctx->t, ctx->color, ctx->games);
 395         }
 396         if (u->pondering) {
 397                 free(ctx->b);
 398                 u->pondering = false;
 399         }
 400 }
 401
 402
 403 void
 404 uct_genmove_setup(struct uct *u, struct board *b, enum stone color)
 405 {
 406         if (b->superko_violation) {
 407                 fprintf(stderr, "!!! WARNING: SUPERKO VIOLATION OCCURED BEFORE THIS MOVE\n");
 408                 fprintf(stderr, "Maybe you play with situational instead of positional superko?\n");
 409                 fprintf(stderr, "I'm going to ignore the violation, but note that I may miss\n");
 410                 fprintf(stderr, "some moves valid under this ruleset because of this.\n");
 411                 b->superko_violation = false;
 412         }
 413
 414         uct_prepare_move(u, b, color);
 415
 416         assert(u->t);
 417         u->my_color = color;
 418
 419         /* How to decide whether to use dynkomi in this game? Since we use
 420          * pondering, it's not simple "who-to-play" matter. Decide based on
 421          * the last genmove issued. */
 422         u->t->use_extra_komi = !!(u->dynkomi_mask & color);
 423 #if 0
 424         /* Moreover, we do not use extra komi at the game end - we are not
 425          * to fool ourselves at this point. */
 426         /* NOTE: But board_estimated_moves_left() is really too rough for this.  On a 19x19 board,
 427          * we would just stop using extra komi after move 210 plus the number of captured stones,
 428          * leading to extremely conservative moves. */
 429         if (board_estimated_moves_left(b) <= MIN_MOVES_LEFT)
 430                 u->t->use_extra_komi = false;
 431 #endif
 432         setup_dynkomi(u, b, color);
 433
 434         if (b->rules == RULES_JAPANESE)
 435                 u->territory_scoring = true;
 436
 437         /* Make pessimistic assumption about komi for Japanese rules to
 438          * avoid losing by 0.5 when winning by 0.5 with Chinese rules.
 439          * The rules usually give the same winner if the integer part of komi
 440          * is odd so we adjust the komi only if it is even (for a board of
 441          * odd size). We are not trying  to get an exact evaluation for rare
 442          * cases of seki. For details see http://home.snafu.de/jasiek/parity.html */
 443         if (u->territory_scoring && (((int)floor(b->komi) + board_size(b)) & 1)) {
 444                 b->komi += (color == S_BLACK ? 1.0 : -1.0);
 445                 if (UDEBUGL(0))
 446                         fprintf(stderr, "Setting komi to %.1f assuming Japanese rules\n",
 447                                 b->komi);
 448         }
 449 }
 450
 451 static coord_t *
 452 uct_genmove(struct engine *e, struct board *b, struct time_info *ti, enum stone color, bool pass_all_alive)
 453 {
 454         double start_time = time_now();
 455         struct uct *u = e->data;
 456         uct_pondering_stop(u);
 457         uct_genmove_setup(u, b, color);
 458
 459         /* Start the Monte Carlo Tree Search! */
 460         int base_playouts = u->t->root->u.playouts;
 461         int played_games = uct_search(u, b, ti, color, u->t);
 462
 463         coord_t best_coord;
 464         struct tree_node *best;
 465         best = uct_search_result(u, b, color, pass_all_alive, played_games, base_playouts, &best_coord);
 466
 467         if (UDEBUGL(2)) {
 468                 double time = time_now() - start_time + 0.000001; /* avoid divide by zero */
 469                 fprintf(stderr, "genmove in %0.2fs (%d games/s, %d games/s/thread)\n",
 470                         time, (int)(played_games/time), (int)(played_games/time/u->threads));
 471         }
 472
 473         if (!best) {
 474                 /* Pass or resign.  Probably not a well-considered pass either. */
 475                 reset_state(u);
 476                 return coord_copy(best_coord);
 477         }
 478         tree_promote_node(u->t, &best);
 479
 480         /* After a pass, pondering is harmful for two reasons:
 481          * (i) We might keep pondering even when the game is over.
 482          * Of course this is the case for opponent resign as well.
 483          * (ii) More importantly, the ownermap will get skewed since
 484          * the UCT will start cutting off any playouts. */
 485         if (u->pondering_opt && !is_pass(best->coord)) {
 486                 uct_pondering_start(u, b, u->t, stone_other(color));
 487         }
 488         return coord_copy(best_coord);
 489 }
 490
 491
 492 bool
 493 uct_gentbook(struct engine *e, struct board *b, struct time_info *ti, enum stone color)
 494 {
 495         struct uct *u = e->data;
 496         if (!u->t) uct_prepare_move(u, b, color);
 497         assert(u->t);
 498
 499         if (ti->dim == TD_GAMES) {
 500                 /* Don't count in games that already went into the tbook. */
 501                 ti->len.games += u->t->root->u.playouts;
 502         }
 503         uct_search(u, b, ti, color, u->t);
 504
 505         assert(ti->dim == TD_GAMES);
 506         tree_save(u->t, b, ti->len.games / 100);
 507
 508         return true;
 509 }
 510
 511 void
 512 uct_dumptbook(struct engine *e, struct board *b, enum stone color)
 513 {
 514         struct uct *u = e->data;
 515         struct tree *t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
 516                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, 0);
 517         tree_load(t, b);
 518         tree_dump(t, 0, true);
 519         tree_done(t);
 520 }
 521
 522
 523 floating_t
 524 uct_evaluate(struct engine *e, struct board *b, struct time_info *ti, coord_t c, enum stone color)
 525 {
 526         struct uct *u = e->data;
 527
 528         struct board b2;
 529         board_copy(&b2, b);
 530         struct move m = { c, color };
 531         int res = board_play(&b2, &m);
 532         if (res < 0)
 533                 return NAN;
 534         color = stone_other(color);
 535
 536         if (u->t) reset_state(u);
 537         uct_prepare_move(u, &b2, color);
 538         assert(u->t);
 539
 540         floating_t bestval;
 541         uct_search(u, &b2, ti, color, u->t);
 542         struct tree_node *best = u->policy->choose(u->policy, u->t->root, &b2, color, resign);
 543         if (!best) {
 544                 bestval = NAN; // the opponent has no reply!
 545         } else {
 546                 bestval = tree_node_get_value(u->t, 1, best->u.value);
 547         }
 548
 549         reset_state(u); // clean our junk
 550
 551         return isnan(bestval) ? NAN : 1.0f - bestval;
 552 }
 553
 554
 555 struct uct *
 556 uct_state_init(char *arg, struct board *b)
 557 {
 558         struct uct *u = calloc2(1, sizeof(struct uct));
 559
 560         u->debug_level = debug_level;
 561         u->gamelen = MC_GAMELEN;
 562         u->resign_threshold = 0.2;
 563         u->sure_win_threshold = 0.85;
 564         u->mercymin = 0;
 565         u->significant_threshold = 50;
 566         u->expand_p = 2;
 567         u->dumpthres = 1000;
 568         u->unlimited_tree_dump = false;
 569         u->playout_amaf = true;
 570         u->playout_amaf_nakade = false;
 571         u->amaf_prior = false;
 572         u->max_tree_size = 1408ULL * 1048576;
 573         u->fast_alloc = true;
 574         u->pruning_threshold = 0;
 575
 576         u->threads = 1;
 577         u->thread_model = TM_TREEVL;
 578         u->virtual_loss = 1;
 579
 580         u->fuseki_end = 20; // max time at 361*20% = 72 moves (our 36th move, still 99 to play)
 581         u->yose_start = 40; // (100-40-25)*361/100/2 = 63 moves still to play by us then
 582         u->bestr_ratio = 0.02;
 583         // 2.5 is clearly too much, but seems to compensate well for overly stern time allocations.
 584         // TODO: Further tuning and experiments with better time allocation schemes.
 585         u->best2_ratio = 2.5;
 586         u->max_maintime_ratio = 8.0;
 587
 588         u->val_scale = 0.04; u->val_points = 40;
 589         u->dynkomi_interval = 1000;
 590         u->dynkomi_mask = S_BLACK | S_WHITE;
 591
 592         u->tenuki_d = 4;
 593         u->local_tree_aging = 80;
 594         u->local_tree_allseq = 1;
 595         u->local_tree_rootseqval = 1;
 596         u->local_tree_depth_decay = 1.5;
 597
 598         u->stats_delay = 0.01; // 10 ms
 599
 600         u->plugins = pluginset_init(b);
 601
 602         u->jdict = joseki_load(b->size);
 603
 604         if (arg) {
 605                 char *optspec, *next = arg;
 606                 while (*next) {
 607                         optspec = next;
 608                         next += strcspn(next, ",");
 609                         if (*next) { *next++ = 0; } else { *next = 0; }
 610
 611                         char *optname = optspec;
 612                         char *optval = strchr(optspec, '=');
 613                         if (optval) *optval++ = 0;
 614
 615                         /** Basic options */
 616
 617                         if (!strcasecmp(optname, "debug")) {
 618                                 if (optval)
 619                                         u->debug_level = atoi(optval);
 620                                 else
 621                                         u->debug_level++;
 622                         } else if (!strcasecmp(optname, "dumpthres") && optval) {
 623                                 /* When dumping the UCT tree on output, include
 624                                  * nodes with at least this many playouts.
 625                                  * (This value is re-scaled "intelligently"
 626                                  * in case of very large trees.) */
 627                                 u->dumpthres = atoi(optval);
 628                         } else if (!strcasecmp(optname, "unlimited_tree_dump")) {
 629                                 /* If true, dumpthres is not increased even if the number of
 630                                  * playouts is much larger than it.  Good for debugging, but could
 631                                  * lead to too much verbosity in actual games. */
 632                                 u->unlimited_tree_dump = true;
 633                         } else if (!strcasecmp(optname, "resign_threshold") && optval) {
 634                                 /* Resign when this ratio of games is lost
 635                                  * after GJ_MINGAMES sample is taken. */
 636                                 u->resign_threshold = atof(optval);
 637                         } else if (!strcasecmp(optname, "sure_win_threshold") && optval) {
 638                                 /* Stop reading when this ratio of games is won
 639                                  * after PLAYOUT_EARLY_BREAK_MIN sample is
 640                                  * taken. (Prevents stupid time losses,
 641                                  * friendly to human opponents.) */
 642                                 u->sure_win_threshold = atof(optval);
 643                         } else if (!strcasecmp(optname, "force_seed") && optval) {
 644                                 /* Set RNG seed at the tree setup. */
 645                                 u->force_seed = atoi(optval);
 646                         } else if (!strcasecmp(optname, "no_tbook")) {
 647                                 /* Disable UCT opening tbook. */
 648                                 u->no_tbook = true;
 649                         } else if (!strcasecmp(optname, "pass_all_alive")) {
 650                                 /* Whether to consider passing only after all
 651                                  * dead groups were removed from the board;
 652                                  * this is like all genmoves are in fact
 653                                  * kgs-genmove_cleanup. */
 654                                 u->pass_all_alive = !optval || atoi(optval);
 655                         } else if (!strcasecmp(optname, "territory_scoring")) {
 656                                 /* Use territory scoring (default is area scoring).
 657                                  * An explicit kgs-rules command overrides this. */
 658                                 u->territory_scoring = !optval || atoi(optval);
 659                         } else if (!strcasecmp(optname, "banner") && optval) {
 660                                 /* Additional banner string. This must come as the
 661                                  * last engine parameter. */
 662                                 if (*next) *--next = ',';
 663                                 u->banner = strdup(optval);
 664                                 break;
 665                         } else if (!strcasecmp(optname, "plugin") && optval) {
 666                                 /* Load an external plugin; filename goes before the colon,
 667                                  * extra arguments after the colon. */
 668                                 char *pluginarg = strchr(optval, ':');
 669                                 if (pluginarg)
 670                                         *pluginarg++ = 0;
 671                                 plugin_load(u->plugins, optval, pluginarg);
 672
 673                         /** UCT behavior and policies */
 674
 675                         } else if ((!strcasecmp(optname, "policy")
 676                                 /* Node selection policy. ucb1amaf is the
 677                                  * default policy implementing RAVE, while
 678                                  * ucb1 is the simple exploration/exploitation
 679                                  * policy. Policies can take further extra
 680                                  * options. */
 681                                     || !strcasecmp(optname, "random_policy")) && optval) {
 682                                 /* A policy to be used randomly with small
 683                                  * chance instead of the default policy. */
 684                                 char *policyarg = strchr(optval, ':');
 685                                 struct uct_policy **p = !strcasecmp(optname, "policy") ? &u->policy : &u->random_policy;
 686                                 if (policyarg)
 687                                         *policyarg++ = 0;
 688                                 if (!strcasecmp(optval, "ucb1")) {
 689                                         *p = policy_ucb1_init(u, policyarg);
 690                                 } else if (!strcasecmp(optval, "ucb1amaf")) {
 691                                         *p = policy_ucb1amaf_init(u, policyarg);
 692                                 } else {
 693                                         fprintf(stderr, "UCT: Invalid tree policy %s\n", optval);
 694                                         exit(1);
 695                                 }
 696                         } else if (!strcasecmp(optname, "playout") && optval) {
 697                                 /* Random simulation (playout) policy.
 698                                  * moggy is the default policy with large
 699                                  * amount of domain-specific knowledge and
 700                                  * heuristics. light is a simple uniformly
 701                                  * random move selection policy. */
 702                                 char *playoutarg = strchr(optval, ':');
 703                                 if (playoutarg)
 704                                         *playoutarg++ = 0;
 705                                 if (!strcasecmp(optval, "moggy")) {
 706                                         u->playout = playout_moggy_init(playoutarg, b, u->jdict);
 707                                 } else if (!strcasecmp(optval, "light")) {
 708                                         u->playout = playout_light_init(playoutarg, b);
 709                                 } else {
 710                                         fprintf(stderr, "UCT: Invalid playout policy %s\n", optval);
 711                                         exit(1);
 712                                 }
 713                         } else if (!strcasecmp(optname, "prior") && optval) {
 714                                 /* Node priors policy. When expanding a node,
 715                                  * it will seed node values heuristically
 716                                  * (most importantly, based on playout policy
 717                                  * opinion, but also with regard to other
 718                                  * things). See uct/prior.c for details.
 719                                  * Use prior=eqex=0 to disable priors. */
 720                                 u->prior = uct_prior_init(optval, b);
 721                         } else if (!strcasecmp(optname, "mercy") && optval) {
 722                                 /* Minimal difference of black/white captures
 723                                  * to stop playout - "Mercy Rule". Speeds up
 724                                  * hopeless playouts at the expense of some
 725                                  * accuracy. */
 726                                 u->mercymin = atoi(optval);
 727                         } else if (!strcasecmp(optname, "gamelen") && optval) {
 728                                 /* Maximum length of single simulation
 729                                  * in moves. */
 730                                 u->gamelen = atoi(optval);
 731                         } else if (!strcasecmp(optname, "expand_p") && optval) {
 732                                 /* Expand UCT nodes after it has been
 733                                  * visited this many times (NOTE: disabled for now). */
 734                                 u->expand_p = atoi(optval);
 735                         } else if (!strcasecmp(optname, "random_policy_chance") && optval) {
 736                                 /* If specified (N), with probability 1/N, random_policy policy
 737                                  * descend is used instead of main policy descend; useful
 738                                  * if specified policy (e.g. UCB1AMAF) can make unduly biased
 739                                  * choices sometimes, you can fall back to e.g.
 740                                  * random_policy=UCB1. */
 741                                 u->random_policy_chance = atoi(optval);
 742
 743                         /** General AMAF behavior */
 744                         /* (Only relevant if the policy supports AMAF.
 745                          * More variables can be tuned as policy
 746                          * parameters.) */
 747
 748                         } else if (!strcasecmp(optname, "playout_amaf")) {
 749                                 /* Whether to include random playout moves in
 750                                  * AMAF as well. (Otherwise, only tree moves
 751                                  * are included in AMAF. Of course makes sense
 752                                  * only in connection with an AMAF policy.) */
 753                                 /* with-without: 55.5% (+-4.1) */
 754                                 if (optval && *optval == '0')
 755                                         u->playout_amaf = false;
 756                                 else
 757                                         u->playout_amaf = true;
 758                         } else if (!strcasecmp(optname, "playout_amaf_nakade")) {
 759                                 /* Whether to include nakade moves from playouts
 760                                  * in the AMAF statistics; this tends to nullify
 761                                  * the playout_amaf effect by adding too much
 762                                  * noise. */
 763                                 if (optval && *optval == '0')
 764                                         u->playout_amaf_nakade = false;
 765                                 else
 766                                         u->playout_amaf_nakade = 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
 869                                          * since this doesn't work well
 870                                          * for white handicaps! */
 871                                         u->dynkomi = uct_dynkomi_init_linear(u, dynkomiarg, b);
 872                                 } else if (!strcasecmp(optval, "adaptive")) {
 873                                         /* There are many more knobs to
 874                                          * crank - see uct/dynkomi.c. */
 875                                         u->dynkomi = uct_dynkomi_init_adaptive(u, dynkomiarg, b);
 876                                 } else {
 877                                         fprintf(stderr, "UCT: Invalid dynkomi mode %s\n", optval);
 878                                         exit(1);
 879                                 }
 880                         } else if (!strcasecmp(optname, "dynkomi_mask") && optval) {
 881                                 /* Bitmask of colors the player must be
 882                                  * for dynkomi be applied; you may want
 883                                  * to use dynkomi_mask=3 to allow dynkomi
 884                                  * even in games where Pachi is white. */
 885                                 u->dynkomi_mask = atoi(optval);
 886                         } else if (!strcasecmp(optname, "dynkomi_interval") && optval) {
 887                                 /* If non-zero, re-adjust dynamic komi
 888                                  * throughout a single genmove reading,
 889                                  * roughly every N simulations. */
 890                                 /* XXX: Does not work with tree
 891                                  * parallelization. */
 892                                 u->dynkomi_interval = atoi(optval);
 893
 894                         /** Node value result scaling */
 895
 896                         } else if (!strcasecmp(optname, "val_scale") && optval) {
 897                                 /* How much of the game result value should be
 898                                  * influenced by win size. Zero means it isn't. */
 899                                 u->val_scale = atof(optval);
 900                         } else if (!strcasecmp(optname, "val_points") && optval) {
 901                                 /* Maximum size of win to be scaled into game
 902                                  * result value. Zero means boardsize^2. */
 903                                 u->val_points = atoi(optval) * 2; // result values are doubled
 904                         } else if (!strcasecmp(optname, "val_extra")) {
 905                                 /* If false, the score coefficient will be simply
 906                                  * added to the value, instead of scaling the result
 907                                  * coefficient because of it. */
 908                                 u->val_extra = !optval || atoi(optval);
 909
 910                         /** Local trees */
 911                         /* (Purely experimental. Does not work - yet!) */
 912
 913                         } else if (!strcasecmp(optname, "local_tree") && optval) {
 914                                 /* Whether to bias exploration by local tree values
 915                                  * (must be supported by the used policy).
 916                                  * 0: Don't.
 917                                  * 1: Do, value = result.
 918                                  * Try to temper the result:
 919                                  * 2: Do, value = 0.5+(result-expected)/2.
 920                                  * 3: Do, value = 0.5+bzz((result-expected)^2).
 921                                  * 4: Do, value = 0.5+sqrt(result-expected)/2. */
 922                                 u->local_tree = atoi(optval);
 923                         } else if (!strcasecmp(optname, "tenuki_d") && optval) {
 924                                 /* Tenuki distance at which to break the local tree. */
 925                                 u->tenuki_d = atoi(optval);
 926                                 if (u->tenuki_d > TREE_NODE_D_MAX + 1) {
 927                                         fprintf(stderr, "uct: tenuki_d must not be larger than TREE_NODE_D_MAX+1 %d\n", TREE_NODE_D_MAX + 1);
 928                                         exit(1);
 929                                 }
 930                         } else if (!strcasecmp(optname, "local_tree_aging") && optval) {
 931                                 /* How much to reduce local tree values between moves. */
 932                                 u->local_tree_aging = atof(optval);
 933                         } else if (!strcasecmp(optname, "local_tree_depth_decay") && optval) {
 934                                 /* With value x>0, during the descent the node
 935                                  * contributes 1/x^depth playouts in
 936                                  * the local tree. I.e., with x>1, nodes more
 937                                  * distant from local situation contribute more
 938                                  * than nodes near the root. */
 939                                 u->local_tree_depth_decay = atof(optval);
 940                         } else if (!strcasecmp(optname, "local_tree_allseq")) {
 941                                 /* If disabled, only complete sequences are stored
 942                                  * in the local tree. If this is on, also
 943                                  * subsequences starting at each move are stored. */
 944                                 u->local_tree_allseq = !optval || atoi(optval);
 945                         } else if (!strcasecmp(optname, "local_tree_rootseqval")) {
 946                                 /* If disabled, expected node value is computed by
 947                                  * summing up values through the whole descent.
 948                                  * If enabled, expected node value for
 949                                  * each sequence is the value at the root of the
 950                                  * sequence. */
 951                                 u->local_tree_rootseqval = !optval || atoi(optval);
 952
 953                         /** Other heuristics */
 954                         } else if (!strcasecmp(optname, "significant_threshold") && optval) {
 955                                 /* Some heuristics (XXX: none in mainline) rely
 956                                  * on the knowledge of the last "significant"
 957                                  * node in the descent. Such a node is
 958                                  * considered reasonably trustworthy to carry
 959                                  * some meaningful information in the values
 960                                  * of the node and its children. */
 961                                 u->significant_threshold = atoi(optval);
 962
 963                         /** Distributed engine slaves setup */
 964
 965                         } else if (!strcasecmp(optname, "slave")) {
 966                                 /* Act as slave for the distributed engine. */
 967                                 u->slave = !optval || atoi(optval);
 968                         } else if (!strcasecmp(optname, "shared_nodes") && optval) {
 969                                 /* Share at most shared_nodes between master and slave at each genmoves.
 970                                  * Must use the same value in master and slaves. */
 971                                 u->shared_nodes = atoi(optval);
 972                         } else if (!strcasecmp(optname, "shared_levels") && optval) {
 973                                 /* Share only nodes of level <= shared_levels. */
 974                                 u->shared_levels = atoi(optval);
 975                         } else if (!strcasecmp(optname, "stats_hbits") && optval) {
 976                                 /* Set hash table size to 2^stats_hbits for the shared stats. */
 977                                 u->stats_hbits = atoi(optval);
 978                         } else if (!strcasecmp(optname, "stats_delay") && optval) {
 979                                 /* How long to wait in slave for initial stats to build up before
 980                                  * replying to the genmoves command (in ms) */
 981                                 u->stats_delay = 0.001 * atof(optval);
 982
 983                         } else {
 984                                 fprintf(stderr, "uct: Invalid engine argument %s or missing value\n", optname);
 985                                 exit(1);
 986                         }
 987                 }
 988         }
 989
 990         if (!u->policy)
 991                 u->policy = policy_ucb1amaf_init(u, NULL);
 992
 993         if (!!u->random_policy_chance ^ !!u->random_policy) {
 994                 fprintf(stderr, "uct: Only one of random_policy and random_policy_chance is set\n");
 995                 exit(1);
 996         }
 997
 998         if (!u->local_tree) {
 999                 /* No ltree aging. */
1000                 u->local_tree_aging = 1.0f;
1001         }
1002
1003         if (u->fast_alloc) {
1004                 if (u->pruning_threshold < u->max_tree_size / 10)
1005                         u->pruning_threshold = u->max_tree_size / 10;
1006                 if (u->pruning_threshold > u->max_tree_size / 2)
1007                         u->pruning_threshold = u->max_tree_size / 2;
1008
1009                 /* Limit pruning temp space to 20% of memory. Beyond this we discard
1010                  * the nodes and recompute them at the next move if necessary. */
1011                 u->max_pruned_size = u->max_tree_size / 5;
1012                 u->max_tree_size -= u->max_pruned_size;
1013         } else {
1014                 /* Reserve 5% memory in case the background free() are slower
1015                  * than the concurrent allocations. */
1016                 u->max_tree_size -= u->max_tree_size / 20;
1017         }
1018
1019         if (!u->prior)
1020                 u->prior = uct_prior_init(NULL, b);
1021
1022         if (!u->playout)
1023                 u->playout = playout_moggy_init(NULL, b, u->jdict);
1024         if (!u->playout->debug_level)
1025                 u->playout->debug_level = u->debug_level;
1026
1027         u->ownermap.map = malloc2(board_size2(b) * sizeof(u->ownermap.map[0]));
1028
1029         if (u->slave) {
1030                 if (!u->stats_hbits) u->stats_hbits = DEFAULT_STATS_HBITS;
1031                 if (!u->shared_nodes) u->shared_nodes = DEFAULT_SHARED_NODES;
1032                 assert(u->shared_levels * board_bits2(b) <= 8 * (int)sizeof(path_t));
1033         }
1034
1035         if (!u->dynkomi)
1036                 u->dynkomi = uct_dynkomi_init_adaptive(u, NULL, b);
1037
1038         /* Some things remain uninitialized for now - the opening tbook
1039          * is not loaded and the tree not set up. */
1040         /* This will be initialized in setup_state() at the first move
1041          * received/requested. This is because right now we are not aware
1042          * about any komi or handicap setup and such. */
1043
1044         return u;
1045 }
1046
1047 struct engine *
1048 engine_uct_init(char *arg, struct board *b)
1049 {
1050         struct uct *u = uct_state_init(arg, b);
1051         struct engine *e = calloc2(1, sizeof(struct engine));
1052         e->name = "UCT Engine";
1053         e->printhook = uct_printhook_ownermap;
1054         e->notify_play = uct_notify_play;
1055         e->chat = uct_chat;
1056         e->undo = uct_undo;
1057         e->result = uct_result;
1058         e->genmove = uct_genmove;
1059         e->genmoves = uct_genmoves;
1060         e->dead_group_list = uct_dead_group_list;
1061         e->done = uct_done;
1062         e->data = u;
1063         if (u->slave)
1064                 e->notify = uct_notify;
1065
1066         const char banner[] = "I'm playing UCT. When I'm losing, I will resign, "
1067                 "if I think I win, I play until you pass. "
1068                 "Anyone can send me 'winrate' in private chat to get my assessment of the position.";
1069         if (!u->banner) u->banner = "";
1070         e->comment = malloc2(sizeof(banner) + strlen(u->banner) + 1);
1071         sprintf(e->comment, "%s %s", banner, u->banner);
1072
1073         return e;
1074 }