uct/uct.c

   1 #include <assert.h>
   2 #include <math.h>
   3 #include <stdio.h>
   4 #include <stdlib.h>
   5 #include <string.h>
   6 #include <time.h>
   7
   8 #define DEBUG
   9
  10 #include "debug.h"
  11 #include "board.h"
  12 #include "gtp.h"
  13 #include "move.h"
  14 #include "mq.h"
  15 #include "joseki/base.h"
  16 #include "playout.h"
  17 #include "playout/moggy.h"
  18 #include "playout/light.h"
  19 #include "tactics/util.h"
  20 #include "timeinfo.h"
  21 #include "uct/dynkomi.h"
  22 #include "uct/internal.h"
  23 #include "uct/plugins.h"
  24 #include "uct/prior.h"
  25 #include "uct/search.h"
  26 #include "uct/slave.h"
  27 #include "uct/tree.h"
  28 #include "uct/uct.h"
  29 #include "uct/walk.h"
  30
  31 struct uct_policy *policy_ucb1_init(struct uct *u, char *arg);
  32 struct uct_policy *policy_ucb1amaf_init(struct uct *u, char *arg, struct board *board);
  33 static void uct_pondering_start(struct uct *u, struct board *b0, struct tree *t, enum stone color);
  34
  35 /* Maximal simulation length. */
  36 #define MC_GAMELEN      MAX_GAMELEN
  37
  38
  39 static void
  40 setup_state(struct uct *u, struct board *b, enum stone color)
  41 {
  42         u->t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
  43                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, u->stats_hbits);
  44         if (u->force_seed)
  45                 fast_srandom(u->force_seed);
  46         if (UDEBUGL(3))
  47                 fprintf(stderr, "Fresh board with random seed %lu\n", fast_getseed());
  48         if (!u->no_tbook && b->moves == 0) {
  49                 if (color == S_BLACK) {
  50                         tree_load(u->t, b);
  51                 } else if (DEBUGL(0)) {
  52                         fprintf(stderr, "Warning: First move appears to be white\n");
  53                 }
  54         }
  55 }
  56
  57 static void
  58 reset_state(struct uct *u)
  59 {
  60         assert(u->t);
  61         tree_done(u->t); u->t = NULL;
  62 }
  63
  64 static void
  65 setup_dynkomi(struct uct *u, struct board *b, enum stone to_play)
  66 {
  67         if (u->t->use_extra_komi && !u->pondering && u->dynkomi->permove)
  68                 u->t->extra_komi = u->dynkomi->permove(u->dynkomi, b, u->t);
  69         else if (!u->t->use_extra_komi)
  70                 u->t->extra_komi = 0;
  71 }
  72
  73 void
  74 uct_prepare_move(struct uct *u, struct board *b, enum stone color)
  75 {
  76         if (u->t) {
  77                 /* Verify that we have sane state. */
  78                 assert(b->es == u);
  79                 assert(u->t && b->moves);
  80                 if (color != stone_other(u->t->root_color)) {
  81                         fprintf(stderr, "Fatal: Non-alternating play detected %d %d\n",
  82                                 color, u->t->root_color);
  83                         exit(1);
  84                 }
  85                 uct_htable_reset(u->t);
  86
  87         } else {
  88                 /* We need fresh state. */
  89                 b->es = u;
  90                 setup_state(u, b, color);
  91         }
  92
  93         u->ownermap.playouts = 0;
  94         memset(u->ownermap.map, 0, board_size2(b) * sizeof(u->ownermap.map[0]));
  95         u->played_own = u->played_all = 0;
  96 }
  97
  98 static void
  99 dead_group_list(struct uct *u, struct board *b, struct move_queue *mq)
 100 {
 101         enum gj_state gs_array[board_size2(b)];
 102         struct group_judgement gj = { .thres = GJ_THRES, .gs = gs_array };
 103         board_ownermap_judge_groups(b, &u->ownermap, &gj);
 104         groups_of_status(b, &gj, GS_DEAD, mq);
 105 }
 106
 107 bool
 108 uct_pass_is_safe(struct uct *u, struct board *b, enum stone color, bool pass_all_alive)
 109 {
 110         /* Make sure enough playouts are simulated to get a reasonable dead group list. */
 111         while (u->ownermap.playouts < GJ_MINGAMES)
 112                 uct_playout(u, b, color, u->t);
 113
 114         struct move_queue mq = { .moves = 0 };
 115         dead_group_list(u, b, &mq);
 116         if (pass_all_alive) {
 117                 for (unsigned int i = 0; i < mq.moves; i++) {
 118                         if (board_at(b, mq.move[i]) == stone_other(color)) {
 119                                 return false; // We need to remove opponent dead groups first.
 120                         }
 121                 }
 122                 mq.moves = 0; // our dead stones are alive when pass_all_alive is true
 123         }
 124         return pass_is_safe(b, color, &mq);
 125 }
 126
 127 static char *
 128 uct_printhook_ownermap(struct board *board, coord_t c, char *s, char *end)
 129 {
 130         struct uct *u = board->es;
 131         if (!u) {
 132                 strcat(s, ". ");
 133                 return s + 2;
 134         }
 135         const char chr[] = ":XO,"; // dame, black, white, unclear
 136         const char chm[] = ":xo,";
 137         char ch = chr[board_ownermap_judge_point(&u->ownermap, c, GJ_THRES)];
 138         if (ch == ',') { // less precise estimate then?
 139                 ch = chm[board_ownermap_judge_point(&u->ownermap, c, 0.67)];
 140         }
 141         s += snprintf(s, end - s, "%c ", ch);
 142         return s;
 143 }
 144
 145 static char *
 146 uct_notify_play(struct engine *e, struct board *b, struct move *m, char *enginearg)
 147 {
 148         struct uct *u = e->data;
 149         if (!u->t) {
 150                 /* No state, create one - this is probably game beginning
 151                  * and we need to load the opening tbook right now. */
 152                 uct_prepare_move(u, b, m->color);
 153                 assert(u->t);
 154         }
 155
 156         /* Stop pondering, required by tree_promote_at() */
 157         uct_pondering_stop(u);
 158         if (UDEBUGL(2) && u->slave)
 159                 tree_dump(u->t, u->dumpthres);
 160
 161         if (is_resign(m->coord)) {
 162                 /* Reset state. */
 163                 reset_state(u);
 164                 return NULL;
 165         }
 166
 167         /* Promote node of the appropriate move to the tree root. */
 168         assert(u->t->root);
 169         if (!tree_promote_at(u->t, b, m->coord)) {
 170                 if (UDEBUGL(3))
 171                         fprintf(stderr, "Warning: Cannot promote move node! Several play commands in row?\n");
 172                 reset_state(u);
 173                 return NULL;
 174         }
 175
 176         /* If we are a slave in a distributed engine, start pondering once
 177          * we know which move we actually played. See uct_genmove() about
 178          * the check for pass. */
 179         if (u->pondering_opt && u->slave && m->color == u->my_color && !is_pass(m->coord))
 180                 uct_pondering_start(u, b, u->t, stone_other(m->color));
 181
 182         return NULL;
 183 }
 184
 185 static char *
 186 uct_undo(struct engine *e, struct board *b)
 187 {
 188         struct uct *u = e->data;
 189
 190         if (!u->t) return NULL;
 191         uct_pondering_stop(u);
 192         reset_state(u);
 193         return NULL;
 194 }
 195
 196 static char *
 197 uct_result(struct engine *e, struct board *b)
 198 {
 199         struct uct *u = e->data;
 200         static char reply[1024];
 201
 202         if (!u->t)
 203                 return NULL;
 204         enum stone color = u->t->root_color;
 205         struct tree_node *n = u->t->root;
 206         snprintf(reply, 1024, "%s %s %d %.2f %.1f",
 207                  stone2str(color), coord2sstr(node_coord(n), b),
 208                  n->u.playouts, tree_node_get_value(u->t, -1, n->u.value),
 209                  u->t->use_extra_komi ? u->t->extra_komi : 0);
 210         return reply;
 211 }
 212
 213 static char *
 214 uct_chat(struct engine *e, struct board *b, char *cmd)
 215 {
 216         struct uct *u = e->data;
 217         static char reply[1024];
 218
 219         cmd += strspn(cmd, " \n\t");
 220         if (!strncasecmp(cmd, "winrate", 7)) {
 221                 if (!u->t)
 222                         return "no game context (yet?)";
 223                 enum stone color = u->t->root_color;
 224                 struct tree_node *n = u->t->root;
 225                 snprintf(reply, 1024, "In %d playouts at %d threads, %s %s can win with %.2f%% probability",
 226                          n->u.playouts, u->threads, stone2str(color), coord2sstr(node_coord(n), b),
 227                          tree_node_get_value(u->t, -1, n->u.value) * 100);
 228                 if (u->t->use_extra_komi && abs(u->t->extra_komi) >= 0.5) {
 229                         sprintf(reply + strlen(reply), ", while self-imposing extra komi %.1f",
 230                                 u->t->extra_komi);
 231                 }
 232                 strcat(reply, ".");
 233                 return reply;
 234         }
 235         return NULL;
 236 }
 237
 238 static void
 239 uct_dead_group_list(struct engine *e, struct board *b, struct move_queue *mq)
 240 {
 241         struct uct *u = e->data;
 242
 243         /* This means the game is probably over, no use pondering on. */
 244         uct_pondering_stop(u);
 245
 246         if (u->pass_all_alive)
 247                 return; // no dead groups
 248
 249         bool mock_state = false;
 250
 251         if (!u->t) {
 252                 /* No state, but we cannot just back out - we might
 253                  * have passed earlier, only assuming some stones are
 254                  * dead, and then re-connected, only to lose counting
 255                  * when all stones are assumed alive. */
 256                 uct_prepare_move(u, b, S_BLACK); assert(u->t);
 257                 mock_state = true;
 258         }
 259         /* Make sure the ownermap is well-seeded. */
 260         while (u->ownermap.playouts < GJ_MINGAMES)
 261                 uct_playout(u, b, S_BLACK, u->t);
 262         /* Show the ownermap: */
 263         if (DEBUGL(2))
 264                 board_print_custom(b, stderr, uct_printhook_ownermap);
 265
 266         dead_group_list(u, b, mq);
 267
 268         if (mock_state) {
 269                 /* Clean up the mock state in case we will receive
 270                  * a genmove; we could get a non-alternating-move
 271                  * error from uct_prepare_move() in that case otherwise. */
 272                 reset_state(u);
 273         }
 274 }
 275
 276 static void
 277 playout_policy_done(struct playout_policy *p)
 278 {
 279         if (p->done) p->done(p);
 280         if (p->data) free(p->data);
 281         free(p);
 282 }
 283
 284 static void
 285 uct_done(struct engine *e)
 286 {
 287         /* This is called on engine reset, especially when clear_board
 288          * is received and new game should begin. */
 289         struct uct *u = e->data;
 290         uct_pondering_stop(u);
 291         if (u->t) reset_state(u);
 292         free(u->ownermap.map);
 293
 294         free(u->policy);
 295         free(u->random_policy);
 296         playout_policy_done(u->playout);
 297         uct_prior_done(u->prior);
 298         joseki_done(u->jdict);
 299         pluginset_done(u->plugins);
 300 }
 301
 302
 303
 304 /* Run time-limited MCTS search on foreground. */
 305 static int
 306 uct_search(struct uct *u, struct board *b, struct time_info *ti, enum stone color, struct tree *t)
 307 {
 308         struct uct_search_state s;
 309         uct_search_start(u, b, color, t, ti, &s);
 310         if (UDEBUGL(2) && s.base_playouts > 0)
 311                 fprintf(stderr, "<pre-simulated %d games>\n", s.base_playouts);
 312
 313         /* The search tree is ctx->t. This is currently == . It is important
 314          * to reference ctx->t directly since the
 315          * thread manager will swap the tree pointer asynchronously. */
 316
 317         /* Now, just periodically poll the search tree. */
 318         /* Note that in case of TD_GAMES, threads will not wait for
 319          * the uct_search_check_stop() signalization. */
 320         while (1) {
 321                 time_sleep(TREE_BUSYWAIT_INTERVAL);
 322                 /* TREE_BUSYWAIT_INTERVAL should never be less than desired time, or the
 323                  * time control is broken. But if it happens to be less, we still search
 324                  * at least 100ms otherwise the move is completely random. */
 325
 326                 int i = uct_search_games(&s);
 327                 /* Print notifications etc. */
 328                 uct_search_progress(u, b, color, t, ti, &s, i);
 329                 /* Check if we should stop the search. */
 330                 if (uct_search_check_stop(u, b, color, t, ti, &s, i))
 331                         break;
 332         }
 333
 334         struct uct_thread_ctx *ctx = uct_search_stop();
 335         if (UDEBUGL(2)) tree_dump(t, u->dumpthres);
 336         if (UDEBUGL(2))
 337                 fprintf(stderr, "(avg score %f/%d; dynkomi's %f/%d value %f/%d)\n",
 338                         t->avg_score.value, t->avg_score.playouts,
 339                         u->dynkomi->score.value, u->dynkomi->score.playouts,
 340                         u->dynkomi->value.value, u->dynkomi->value.playouts);
 341         uct_progress_status(u, t, color, ctx->games, true);
 342
 343         u->played_own += ctx->games;
 344         return ctx->games;
 345 }
 346
 347 /* Start pondering background with @color to play. */
 348 static void
 349 uct_pondering_start(struct uct *u, struct board *b0, struct tree *t, enum stone color)
 350 {
 351         if (UDEBUGL(1))
 352                 fprintf(stderr, "Starting to ponder with color %s\n", stone2str(stone_other(color)));
 353         u->pondering = true;
 354
 355         /* We need a local board copy to ponder upon. */
 356         struct board *b = malloc2(sizeof(*b)); board_copy(b, b0);
 357
 358         /* *b0 did not have the genmove'd move played yet. */
 359         struct move m = { node_coord(t->root), t->root_color };
 360         int res = board_play(b, &m);
 361         assert(res >= 0);
 362         setup_dynkomi(u, b, stone_other(m.color));
 363
 364         /* Start MCTS manager thread "headless". */
 365         static struct uct_search_state s;
 366         uct_search_start(u, b, color, t, NULL, &s);
 367 }
 368
 369 /* uct_search_stop() frontend for the pondering (non-genmove) mode, and
 370  * to stop the background search for a slave in the distributed engine. */
 371 void
 372 uct_pondering_stop(struct uct *u)
 373 {
 374         if (!thread_manager_running)
 375                 return;
 376
 377         /* Stop the thread manager. */
 378         struct uct_thread_ctx *ctx = uct_search_stop();
 379         if (UDEBUGL(1)) {
 380                 if (u->pondering) fprintf(stderr, "(pondering) ");
 381                 uct_progress_status(u, ctx->t, ctx->color, ctx->games, true);
 382         }
 383         if (u->pondering) {
 384                 free(ctx->b);
 385                 u->pondering = false;
 386         }
 387 }
 388
 389
 390 void
 391 uct_genmove_setup(struct uct *u, struct board *b, enum stone color)
 392 {
 393         if (b->superko_violation) {
 394                 fprintf(stderr, "!!! WARNING: SUPERKO VIOLATION OCCURED BEFORE THIS MOVE\n");
 395                 fprintf(stderr, "Maybe you play with situational instead of positional superko?\n");
 396                 fprintf(stderr, "I'm going to ignore the violation, but note that I may miss\n");
 397                 fprintf(stderr, "some moves valid under this ruleset because of this.\n");
 398                 b->superko_violation = false;
 399         }
 400
 401         uct_prepare_move(u, b, color);
 402
 403         assert(u->t);
 404         u->my_color = color;
 405
 406         /* How to decide whether to use dynkomi in this game? Since we use
 407          * pondering, it's not simple "who-to-play" matter. Decide based on
 408          * the last genmove issued. */
 409         u->t->use_extra_komi = !!(u->dynkomi_mask & color);
 410         setup_dynkomi(u, b, color);
 411
 412         if (b->rules == RULES_JAPANESE)
 413                 u->territory_scoring = true;
 414
 415         /* Make pessimistic assumption about komi for Japanese rules to
 416          * avoid losing by 0.5 when winning by 0.5 with Chinese rules.
 417          * The rules usually give the same winner if the integer part of komi
 418          * is odd so we adjust the komi only if it is even (for a board of
 419          * odd size). We are not trying  to get an exact evaluation for rare
 420          * cases of seki. For details see http://home.snafu.de/jasiek/parity.html */
 421         if (u->territory_scoring && (((int)floor(b->komi) + board_size(b)) & 1)) {
 422                 b->komi += (color == S_BLACK ? 1.0 : -1.0);
 423                 if (UDEBUGL(0))
 424                         fprintf(stderr, "Setting komi to %.1f assuming Japanese rules\n",
 425                                 b->komi);
 426         }
 427 }
 428
 429 static coord_t *
 430 uct_genmove(struct engine *e, struct board *b, struct time_info *ti, enum stone color, bool pass_all_alive)
 431 {
 432         double start_time = time_now();
 433         struct uct *u = e->data;
 434         u->pass_all_alive |= pass_all_alive;
 435         uct_pondering_stop(u);
 436         uct_genmove_setup(u, b, color);
 437
 438         /* Start the Monte Carlo Tree Search! */
 439         int base_playouts = u->t->root->u.playouts;
 440         int played_games = uct_search(u, b, ti, color, u->t);
 441
 442         coord_t best_coord;
 443         struct tree_node *best;
 444         best = uct_search_result(u, b, color, u->pass_all_alive, played_games, base_playouts, &best_coord);
 445
 446         if (UDEBUGL(2)) {
 447                 double time = time_now() - start_time + 0.000001; /* avoid divide by zero */
 448                 fprintf(stderr, "genmove in %0.2fs (%d games/s, %d games/s/thread)\n",
 449                         time, (int)(played_games/time), (int)(played_games/time/u->threads));
 450         }
 451
 452         if (!best) {
 453                 /* Pass or resign. */
 454                 reset_state(u);
 455                 return coord_copy(best_coord);
 456         }
 457         tree_promote_node(u->t, &best);
 458
 459         /* After a pass, pondering is harmful for two reasons:
 460          * (i) We might keep pondering even when the game is over.
 461          * Of course this is the case for opponent resign as well.
 462          * (ii) More importantly, the ownermap will get skewed since
 463          * the UCT will start cutting off any playouts. */
 464         if (u->pondering_opt && !is_pass(node_coord(best))) {
 465                 uct_pondering_start(u, b, u->t, stone_other(color));
 466         }
 467         return coord_copy(best_coord);
 468 }
 469
 470
 471 bool
 472 uct_gentbook(struct engine *e, struct board *b, struct time_info *ti, enum stone color)
 473 {
 474         struct uct *u = e->data;
 475         if (!u->t) uct_prepare_move(u, b, color);
 476         assert(u->t);
 477
 478         if (ti->dim == TD_GAMES) {
 479                 /* Don't count in games that already went into the tbook. */
 480                 ti->len.games += u->t->root->u.playouts;
 481         }
 482         uct_search(u, b, ti, color, u->t);
 483
 484         assert(ti->dim == TD_GAMES);
 485         tree_save(u->t, b, ti->len.games / 100);
 486
 487         return true;
 488 }
 489
 490 void
 491 uct_dumptbook(struct engine *e, struct board *b, enum stone color)
 492 {
 493         struct uct *u = e->data;
 494         struct tree *t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
 495                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, 0);
 496         tree_load(t, b);
 497         tree_dump(t, 0);
 498         tree_done(t);
 499 }
 500
 501
 502 floating_t
 503 uct_evaluate_one(struct engine *e, struct board *b, struct time_info *ti, coord_t c, enum stone color)
 504 {
 505         struct uct *u = e->data;
 506
 507         struct board b2;
 508         board_copy(&b2, b);
 509         struct move m = { c, color };
 510         int res = board_play(&b2, &m);
 511         if (res < 0)
 512                 return NAN;
 513         color = stone_other(color);
 514
 515         if (u->t) reset_state(u);
 516         uct_prepare_move(u, &b2, color);
 517         assert(u->t);
 518
 519         floating_t bestval;
 520         uct_search(u, &b2, ti, color, u->t);
 521         struct tree_node *best = u->policy->choose(u->policy, u->t->root, &b2, color, resign);
 522         if (!best) {
 523                 bestval = NAN; // the opponent has no reply!
 524         } else {
 525                 bestval = tree_node_get_value(u->t, 1, best->u.value);
 526         }
 527
 528         reset_state(u); // clean our junk
 529
 530         return isnan(bestval) ? NAN : 1.0f - bestval;
 531 }
 532
 533 void
 534 uct_evaluate(struct engine *e, struct board *b, struct time_info *ti, floating_t *vals, enum stone color)
 535 {
 536         for (int i = 0; i < b->flen; i++) {
 537                 if (is_pass(b->f[i]))
 538                         vals[i] = NAN;
 539                 else
 540                         vals[i] = uct_evaluate_one(e, b, ti, b->f[i], color);
 541         }
 542 }
 543
 544
 545 struct uct *
 546 uct_state_init(char *arg, struct board *b)
 547 {
 548         struct uct *u = calloc2(1, sizeof(struct uct));
 549         bool pat_setup = false;
 550
 551         u->debug_level = debug_level;
 552         u->reportfreq = 10000;
 553         u->gamelen = MC_GAMELEN;
 554         u->resign_threshold = 0.2;
 555         u->sure_win_threshold = 0.9;
 556         u->mercymin = 0;
 557         u->significant_threshold = 50;
 558         u->expand_p = 8;
 559         u->dumpthres = 1000;
 560         u->playout_amaf = true;
 561         u->amaf_prior = false;
 562         u->max_tree_size = 1408ULL * 1048576;
 563         u->fast_alloc = true;
 564         u->pruning_threshold = 0;
 565
 566         u->threads = 1;
 567         u->thread_model = TM_TREEVL;
 568         u->virtual_loss = 1;
 569
 570         u->fuseki_end = 20; // max time at 361*20% = 72 moves (our 36th move, still 99 to play)
 571         u->yose_start = 40; // (100-40-25)*361/100/2 = 63 moves still to play by us then
 572         u->bestr_ratio = 0.02;
 573         // 2.5 is clearly too much, but seems to compensate well for overly stern time allocations.
 574         // TODO: Further tuning and experiments with better time allocation schemes.
 575         u->best2_ratio = 2.5;
 576         u->max_maintime_ratio = 3.0;
 577
 578         u->val_scale = 0; u->val_points = 40;
 579         u->dynkomi_interval = 1000;
 580         u->dynkomi_mask = S_BLACK | S_WHITE;
 581
 582         u->tenuki_d = 4;
 583         u->local_tree_aging = 80;
 584         u->local_tree_depth_decay = 1.5;
 585         u->local_tree_eval = LTE_ROOT;
 586         u->local_tree_neival = true;
 587
 588         u->max_slaves = -1;
 589         u->slave_index = -1;
 590         u->stats_delay = 0.01; // 10 ms
 591
 592         u->plugins = pluginset_init(b);
 593
 594         u->jdict = joseki_load(b->size);
 595
 596         if (arg) {
 597                 char *optspec, *next = arg;
 598                 while (*next) {
 599                         optspec = next;
 600                         next += strcspn(next, ",");
 601                         if (*next) { *next++ = 0; } else { *next = 0; }
 602
 603                         char *optname = optspec;
 604                         char *optval = strchr(optspec, '=');
 605                         if (optval) *optval++ = 0;
 606
 607                         /** Basic options */
 608
 609                         if (!strcasecmp(optname, "debug")) {
 610                                 if (optval)
 611                                         u->debug_level = atoi(optval);
 612                                 else
 613                                         u->debug_level++;
 614                         } else if (!strcasecmp(optname, "reporting") && optval) {
 615                                 /* The format of output for detailed progress
 616                                  * information (such as current best move and
 617                                  * its value, etc.). */
 618                                 if (!strcasecmp(optval, "text")) {
 619                                         /* Plaintext traditional output. */
 620                                         u->reporting = UR_TEXT;
 621                                 } else if (!strcasecmp(optval, "json")) {
 622                                         /* JSON output. Implies debug=0. */
 623                                         u->reporting = UR_JSON;
 624                                         u->debug_level = 0;
 625                                 } else if (!strcasecmp(optval, "jsonbig")) {
 626                                         /* JSON output, but much more detailed.
 627                                          * Implies debug=0. */
 628                                         u->reporting = UR_JSON_BIG;
 629                                         u->debug_level = 0;
 630                                 } else {
 631                                         fprintf(stderr, "UCT: Invalid reporting format %s\n", optval);
 632                                         exit(1);
 633                                 }
 634                         } else if (!strcasecmp(optname, "reportfreq") && optval) {
 635                                 /* The progress information line will be shown
 636                                  * every <reportfreq> simulations. */
 637                                 u->reportfreq = atoi(optval);
 638                         } else if (!strcasecmp(optname, "dumpthres") && optval) {
 639                                 /* When dumping the UCT tree on output, include
 640                                  * nodes with at least this many playouts.
 641                                  * (This value is re-scaled "intelligently"
 642                                  * in case of very large trees.) */
 643                                 u->dumpthres = atoi(optval);
 644                         } else if (!strcasecmp(optname, "resign_threshold") && optval) {
 645                                 /* Resign when this ratio of games is lost
 646                                  * after GJ_MINGAMES sample is taken. */
 647                                 u->resign_threshold = atof(optval);
 648                         } else if (!strcasecmp(optname, "sure_win_threshold") && optval) {
 649                                 /* Stop reading when this ratio of games is won
 650                                  * after PLAYOUT_EARLY_BREAK_MIN sample is
 651                                  * taken. (Prevents stupid time losses,
 652                                  * friendly to human opponents.) */
 653                                 u->sure_win_threshold = atof(optval);
 654                         } else if (!strcasecmp(optname, "force_seed") && optval) {
 655                                 /* Set RNG seed at the tree setup. */
 656                                 u->force_seed = atoi(optval);
 657                         } else if (!strcasecmp(optname, "no_tbook")) {
 658                                 /* Disable UCT opening tbook. */
 659                                 u->no_tbook = true;
 660                         } else if (!strcasecmp(optname, "pass_all_alive")) {
 661                                 /* Whether to consider passing only after all
 662                                  * dead groups were removed from the board;
 663                                  * this is like all genmoves are in fact
 664                                  * kgs-genmove_cleanup. */
 665                                 u->pass_all_alive = !optval || atoi(optval);
 666                         } else if (!strcasecmp(optname, "territory_scoring")) {
 667                                 /* Use territory scoring (default is area scoring).
 668                                  * An explicit kgs-rules command overrides this. */
 669                                 u->territory_scoring = !optval || atoi(optval);
 670                         } else if (!strcasecmp(optname, "stones_only")) {
 671                                 /* Do not count eyes. Nice to teach go to kids.
 672                                  * http://strasbourg.jeudego.org/regle_strasbourgeoise.htm */
 673                                 b->rules = RULES_STONES_ONLY;
 674                                 u->pass_all_alive = true;
 675                         } else if (!strcasecmp(optname, "banner") && optval) {
 676                                 /* Additional banner string. This must come as the
 677                                  * last engine parameter. */
 678                                 if (*next) *--next = ',';
 679                                 u->banner = strdup(optval);
 680                                 break;
 681                         } else if (!strcasecmp(optname, "plugin") && optval) {
 682                                 /* Load an external plugin; filename goes before the colon,
 683                                  * extra arguments after the colon. */
 684                                 char *pluginarg = strchr(optval, ':');
 685                                 if (pluginarg)
 686                                         *pluginarg++ = 0;
 687                                 plugin_load(u->plugins, optval, pluginarg);
 688
 689                         /** UCT behavior and policies */
 690
 691                         } else if ((!strcasecmp(optname, "policy")
 692                                 /* Node selection policy. ucb1amaf is the
 693                                  * default policy implementing RAVE, while
 694                                  * ucb1 is the simple exploration/exploitation
 695                                  * policy. Policies can take further extra
 696                                  * options. */
 697                                     || !strcasecmp(optname, "random_policy")) && optval) {
 698                                 /* A policy to be used randomly with small
 699                                  * chance instead of the default policy. */
 700                                 char *policyarg = strchr(optval, ':');
 701                                 struct uct_policy **p = !strcasecmp(optname, "policy") ? &u->policy : &u->random_policy;
 702                                 if (policyarg)
 703                                         *policyarg++ = 0;
 704                                 if (!strcasecmp(optval, "ucb1")) {
 705                                         *p = policy_ucb1_init(u, policyarg);
 706                                 } else if (!strcasecmp(optval, "ucb1amaf")) {
 707                                         *p = policy_ucb1amaf_init(u, policyarg, b);
 708                                 } else {
 709                                         fprintf(stderr, "UCT: Invalid tree policy %s\n", optval);
 710                                         exit(1);
 711                                 }
 712                         } else if (!strcasecmp(optname, "playout") && optval) {
 713                                 /* Random simulation (playout) policy.
 714                                  * moggy is the default policy with large
 715                                  * amount of domain-specific knowledge and
 716                                  * heuristics. light is a simple uniformly
 717                                  * random move selection policy. */
 718                                 char *playoutarg = strchr(optval, ':');
 719                                 if (playoutarg)
 720                                         *playoutarg++ = 0;
 721                                 if (!strcasecmp(optval, "moggy")) {
 722                                         u->playout = playout_moggy_init(playoutarg, b, u->jdict);
 723                                 } else if (!strcasecmp(optval, "light")) {
 724                                         u->playout = playout_light_init(playoutarg, b);
 725                                 } else {
 726                                         fprintf(stderr, "UCT: Invalid playout policy %s\n", optval);
 727                                         exit(1);
 728                                 }
 729                         } else if (!strcasecmp(optname, "prior") && optval) {
 730                                 /* Node priors policy. When expanding a node,
 731                                  * it will seed node values heuristically
 732                                  * (most importantly, based on playout policy
 733                                  * opinion, but also with regard to other
 734                                  * things). See uct/prior.c for details.
 735                                  * Use prior=eqex=0 to disable priors. */
 736                                 u->prior = uct_prior_init(optval, b, u);
 737                         } else if (!strcasecmp(optname, "mercy") && optval) {
 738                                 /* Minimal difference of black/white captures
 739                                  * to stop playout - "Mercy Rule". Speeds up
 740                                  * hopeless playouts at the expense of some
 741                                  * accuracy. */
 742                                 u->mercymin = atoi(optval);
 743                         } else if (!strcasecmp(optname, "gamelen") && optval) {
 744                                 /* Maximum length of single simulation
 745                                  * in moves. */
 746                                 u->gamelen = atoi(optval);
 747                         } else if (!strcasecmp(optname, "expand_p") && optval) {
 748                                 /* Expand UCT nodes after it has been
 749                                  * visited this many times. */
 750                                 u->expand_p = atoi(optval);
 751                         } else if (!strcasecmp(optname, "random_policy_chance") && optval) {
 752                                 /* If specified (N), with probability 1/N, random_policy policy
 753                                  * descend is used instead of main policy descend; useful
 754                                  * if specified policy (e.g. UCB1AMAF) can make unduly biased
 755                                  * choices sometimes, you can fall back to e.g.
 756                                  * random_policy=UCB1. */
 757                                 u->random_policy_chance = atoi(optval);
 758
 759                         /** General AMAF behavior */
 760                         /* (Only relevant if the policy supports AMAF.
 761                          * More variables can be tuned as policy
 762                          * parameters.) */
 763
 764                         } else if (!strcasecmp(optname, "playout_amaf")) {
 765                                 /* Whether to include random playout moves in
 766                                  * AMAF as well. (Otherwise, only tree moves
 767                                  * are included in AMAF. Of course makes sense
 768                                  * only in connection with an AMAF policy.) */
 769                                 /* with-without: 55.5% (+-4.1) */
 770                                 if (optval && *optval == '0')
 771                                         u->playout_amaf = false;
 772                                 else
 773                                         u->playout_amaf = true;
 774                         } else if (!strcasecmp(optname, "playout_amaf_cutoff") && optval) {
 775                                 /* Keep only first N% of playout stage AMAF
 776                                  * information. */
 777                                 u->playout_amaf_cutoff = atoi(optval);
 778                         } else if (!strcasecmp(optname, "amaf_prior") && optval) {
 779                                 /* In node policy, consider prior values
 780                                  * part of the real result term or part
 781                                  * of the AMAF term? */
 782                                 u->amaf_prior = atoi(optval);
 783
 784                         /** Performance and memory management */
 785
 786                         } else if (!strcasecmp(optname, "threads") && optval) {
 787                                 /* By default, Pachi will run with only single
 788                                  * tree search thread! */
 789                                 u->threads = atoi(optval);
 790                         } else if (!strcasecmp(optname, "thread_model") && optval) {
 791                                 if (!strcasecmp(optval, "tree")) {
 792                                         /* Tree parallelization - all threads
 793                                          * grind on the same tree. */
 794                                         u->thread_model = TM_TREE;
 795                                         u->virtual_loss = 0;
 796                                 } else if (!strcasecmp(optval, "treevl")) {
 797                                         /* Tree parallelization, but also
 798                                          * with virtual losses - this discou-
 799                                          * rages most threads choosing the
 800                                          * same tree branches to read. */
 801                                         u->thread_model = TM_TREEVL;
 802                                 } else {
 803                                         fprintf(stderr, "UCT: Invalid thread model %s\n", optval);
 804                                         exit(1);
 805                                 }
 806                         } else if (!strcasecmp(optname, "virtual_loss")) {
 807                                 /* Number of virtual losses added before evaluating a node. */
 808                                 u->virtual_loss = !optval || atoi(optval);
 809                         } else if (!strcasecmp(optname, "pondering")) {
 810                                 /* Keep searching even during opponent's turn. */
 811                                 u->pondering_opt = !optval || atoi(optval);
 812                         } else if (!strcasecmp(optname, "max_tree_size") && optval) {
 813                                 /* Maximum amount of memory [MiB] consumed by the move tree.
 814                                  * For fast_alloc it includes the temp tree used for pruning.
 815                                  * Default is 3072 (3 GiB). */
 816                                 u->max_tree_size = atol(optval) * 1048576;
 817                         } else if (!strcasecmp(optname, "fast_alloc")) {
 818                                 u->fast_alloc = !optval || atoi(optval);
 819                         } else if (!strcasecmp(optname, "pruning_threshold") && optval) {
 820                                 /* Force pruning at beginning of a move if the tree consumes
 821                                  * more than this [MiB]. Default is 10% of max_tree_size.
 822                                  * Increase to reduce pruning time overhead if memory is plentiful.
 823                                  * This option is meaningful only for fast_alloc. */
 824                                 u->pruning_threshold = atol(optval) * 1048576;
 825
 826                         /** Time control */
 827
 828                         } else if (!strcasecmp(optname, "best2_ratio") && optval) {
 829                                 /* If set, prolong simulating while
 830                                  * first_best/second_best playouts ratio
 831                                  * is less than best2_ratio. */
 832                                 u->best2_ratio = atof(optval);
 833                         } else if (!strcasecmp(optname, "bestr_ratio") && optval) {
 834                                 /* If set, prolong simulating while
 835                                  * best,best_best_child values delta
 836                                  * is more than bestr_ratio. */
 837                                 u->bestr_ratio = atof(optval);
 838                         } else if (!strcasecmp(optname, "max_maintime_ratio") && optval) {
 839                                 /* If set and while not in byoyomi, prolong simulating no more than
 840                                  * max_maintime_ratio times the normal desired thinking time. */
 841                                 u->max_maintime_ratio = atof(optval);
 842                         } else if (!strcasecmp(optname, "fuseki_end") && optval) {
 843                                 /* At the very beginning it's not worth thinking
 844                                  * too long because the playout evaluations are
 845                                  * very noisy. So gradually increase the thinking
 846                                  * time up to maximum when fuseki_end percent
 847                                  * of the board has been played.
 848                                  * This only applies if we are not in byoyomi. */
 849                                 u->fuseki_end = atoi(optval);
 850                         } else if (!strcasecmp(optname, "yose_start") && optval) {
 851                                 /* When yose_start percent of the board has been
 852                                  * played, or if we are in byoyomi, stop spending
 853                                  * more time and spread the remaining time
 854                                  * uniformly.
 855                                  * Between fuseki_end and yose_start, we spend
 856                                  * a constant proportion of the remaining time
 857                                  * on each move. (yose_start should actually
 858                                  * be much earlier than when real yose start,
 859                                  * but "yose" is a good short name to convey
 860                                  * the idea.) */
 861                                 u->yose_start = atoi(optval);
 862
 863                         /** Dynamic komi */
 864
 865                         } else if (!strcasecmp(optname, "dynkomi") && optval) {
 866                                 /* Dynamic komi approach; there are multiple
 867                                  * ways to adjust komi dynamically throughout
 868                                  * play. We currently support two: */
 869                                 char *dynkomiarg = strchr(optval, ':');
 870                                 if (dynkomiarg)
 871                                         *dynkomiarg++ = 0;
 872                                 if (!strcasecmp(optval, "none")) {
 873                                         u->dynkomi = uct_dynkomi_init_none(u, dynkomiarg, b);
 874                                 } else if (!strcasecmp(optval, "linear")) {
 875                                         /* You should set dynkomi_mask=1 or a very low
 876                                          * handicap_value for white. */
 877                                         u->dynkomi = uct_dynkomi_init_linear(u, dynkomiarg, b);
 878                                 } else if (!strcasecmp(optval, "adaptive")) {
 879                                         /* There are many more knobs to
 880                                          * crank - see uct/dynkomi.c. */
 881                                         u->dynkomi = uct_dynkomi_init_adaptive(u, dynkomiarg, b);
 882                                 } else {
 883                                         fprintf(stderr, "UCT: Invalid dynkomi mode %s\n", optval);
 884                                         exit(1);
 885                                 }
 886                         } else if (!strcasecmp(optname, "dynkomi_mask") && optval) {
 887                                 /* Bitmask of colors the player must be
 888                                  * for dynkomi be applied; the default dynkomi_mask=3 allows
 889                                  * dynkomi even in games where Pachi is white. */
 890                                 u->dynkomi_mask = atoi(optval);
 891                         } else if (!strcasecmp(optname, "dynkomi_interval") && optval) {
 892                                 /* If non-zero, re-adjust dynamic komi
 893                                  * throughout a single genmove reading,
 894                                  * roughly every N simulations. */
 895                                 /* XXX: Does not work with tree
 896                                  * parallelization. */
 897                                 u->dynkomi_interval = atoi(optval);
 898
 899                         /** Node value result scaling */
 900
 901                         } else if (!strcasecmp(optname, "val_scale") && optval) {
 902                                 /* How much of the game result value should be
 903                                  * influenced by win size. Zero means it isn't. */
 904                                 u->val_scale = atof(optval);
 905                         } else if (!strcasecmp(optname, "val_points") && optval) {
 906                                 /* Maximum size of win to be scaled into game
 907                                  * result value. Zero means boardsize^2. */
 908                                 u->val_points = atoi(optval) * 2; // result values are doubled
 909                         } else if (!strcasecmp(optname, "val_extra")) {
 910                                 /* If false, the score coefficient will be simply
 911                                  * added to the value, instead of scaling the result
 912                                  * coefficient because of it. */
 913                                 u->val_extra = !optval || atoi(optval);
 914
 915                         /** Local trees */
 916                         /* (Purely experimental. Does not work - yet!) */
 917
 918                         } else if (!strcasecmp(optname, "local_tree")) {
 919                                 /* Whether to bias exploration by local tree values. */
 920                                 u->local_tree = !optval || atoi(optval);
 921                         } else if (!strcasecmp(optname, "tenuki_d") && optval) {
 922                                 /* Tenuki distance at which to break the local tree. */
 923                                 u->tenuki_d = atoi(optval);
 924                                 if (u->tenuki_d > TREE_NODE_D_MAX + 1) {
 925                                         fprintf(stderr, "uct: tenuki_d must not be larger than TREE_NODE_D_MAX+1 %d\n", TREE_NODE_D_MAX + 1);
 926                                         exit(1);
 927                                 }
 928                         } else if (!strcasecmp(optname, "local_tree_aging") && optval) {
 929                                 /* How much to reduce local tree values between moves. */
 930                                 u->local_tree_aging = atof(optval);
 931                         } else if (!strcasecmp(optname, "local_tree_depth_decay") && optval) {
 932                                 /* With value x>0, during the descent the node
 933                                  * contributes 1/x^depth playouts in
 934                                  * the local tree. I.e., with x>1, nodes more
 935                                  * distant from local situation contribute more
 936                                  * than nodes near the root. */
 937                                 u->local_tree_depth_decay = atof(optval);
 938                         } else if (!strcasecmp(optname, "local_tree_allseq")) {
 939                                 /* If disabled, only complete sequences are stored
 940                                  * in the local tree. If this is on, also
 941                                  * subsequences starting at each move are stored. */
 942                                 u->local_tree_allseq = !optval || atoi(optval);
 943                         } else if (!strcasecmp(optname, "local_tree_neival")) {
 944                                 /* If disabled, local node value is not
 945                                  * computed just based on terminal status
 946                                  * of the coordinate, but also its neighbors. */
 947                                 u->local_tree_neival = !optval || atoi(optval);
 948                         } else if (!strcasecmp(optname, "local_tree_eval")) {
 949                                 /* How is the value inserted in the local tree
 950                                  * determined. */
 951                                 if (!strcasecmp(optval, "root"))
 952                                         /* All moves within a tree branch are
 953                                          * considered wrt. their merit
 954                                          * reaching tachtical goal of making
 955                                          * the first move in the branch
 956                                          * survive. */
 957                                         u->local_tree_eval = LTE_ROOT;
 958                                 else if (!strcasecmp(optval, "each"))
 959                                         /* Each move is considered wrt.
 960                                          * its own survival. */
 961                                         u->local_tree_eval = LTE_EACH;
 962                                 else if (!strcasecmp(optval, "total"))
 963                                         /* The tactical goal is the survival
 964                                          * of all the moves of my color and
 965                                          * non-survival of all the opponent
 966                                          * moves. Local values (and their
 967                                          * inverses) are averaged. */
 968                                         u->local_tree_eval = LTE_TOTAL;
 969                                 else {
 970                                         fprintf(stderr, "uct: unknown local_tree_eval %s\n", optval);
 971                                         exit(1);
 972                                 }
 973                         } else if (!strcasecmp(optname, "local_tree_rootchoose")) {
 974                                 /* If disabled, only moves within the local
 975                                  * tree branch are considered; the values
 976                                  * of the branch roots (i.e. root children)
 977                                  * are ignored. This may make sense together
 978                                  * with eval!=each, we consider only moves
 979                                  * that influence the goal, not the "rating"
 980                                  * of the goal itself. (The real solution
 981                                  * will be probably using criticality to pick
 982                                  * local tree branches.) */
 983                                 u->local_tree_rootchoose = !optval || atoi(optval);
 984
 985                         /** Other heuristics */
 986                         } else if (!strcasecmp(optname, "patterns")) {
 987                                 /* Load pattern database. Various modules
 988                                  * (priors, policies etc.) may make use
 989                                  * of this database. They will request
 990                                  * it automatically in that case, but you
 991                                  * can use this option to tweak the pattern
 992                                  * parameters. */
 993                                 patterns_init(&u->pat, optval, false, true);
 994                                 u->want_pat = pat_setup = true;
 995                         } else if (!strcasecmp(optname, "significant_threshold") && optval) {
 996                                 /* Some heuristics (XXX: none in mainline) rely
 997                                  * on the knowledge of the last "significant"
 998                                  * node in the descent. Such a node is
 999                                  * considered reasonably trustworthy to carry
1000                                  * some meaningful information in the values
1001                                  * of the node and its children. */
1002                                 u->significant_threshold = atoi(optval);
1003
1004                         /** Distributed engine slaves setup */
1005
1006                         } else if (!strcasecmp(optname, "slave")) {
1007                                 /* Act as slave for the distributed engine. */
1008                                 u->slave = !optval || atoi(optval);
1009                         } else if (!strcasecmp(optname, "slave_index") && optval) {
1010                                 /* Optional index if per-slave behavior is desired.
1011                                  * Must be given as index/max */
1012                                 u->slave_index = atoi(optval);
1013                                 char *p = strchr(optval, '/');
1014                                 if (p) u->max_slaves = atoi(++p);
1015                         } else if (!strcasecmp(optname, "shared_nodes") && optval) {
1016                                 /* Share at most shared_nodes between master and slave at each genmoves.
1017                                  * Must use the same value in master and slaves. */
1018                                 u->shared_nodes = atoi(optval);
1019                         } else if (!strcasecmp(optname, "shared_levels") && optval) {
1020                                 /* Share only nodes of level <= shared_levels. */
1021                                 u->shared_levels = atoi(optval);
1022                         } else if (!strcasecmp(optname, "stats_hbits") && optval) {
1023                                 /* Set hash table size to 2^stats_hbits for the shared stats. */
1024                                 u->stats_hbits = atoi(optval);
1025                         } else if (!strcasecmp(optname, "stats_delay") && optval) {
1026                                 /* How long to wait in slave for initial stats to build up before
1027                                  * replying to the genmoves command (in ms) */
1028                                 u->stats_delay = 0.001 * atof(optval);
1029
1030                         } else {
1031                                 fprintf(stderr, "uct: Invalid engine argument %s or missing value\n", optname);
1032                                 exit(1);
1033                         }
1034                 }
1035         }
1036
1037         if (!u->policy)
1038                 u->policy = policy_ucb1amaf_init(u, NULL, b);
1039
1040         if (!!u->random_policy_chance ^ !!u->random_policy) {
1041                 fprintf(stderr, "uct: Only one of random_policy and random_policy_chance is set\n");
1042                 exit(1);
1043         }
1044
1045         if (!u->local_tree) {
1046                 /* No ltree aging. */
1047                 u->local_tree_aging = 1.0f;
1048         }
1049
1050         if (u->fast_alloc) {
1051                 if (u->pruning_threshold < u->max_tree_size / 10)
1052                         u->pruning_threshold = u->max_tree_size / 10;
1053                 if (u->pruning_threshold > u->max_tree_size / 2)
1054                         u->pruning_threshold = u->max_tree_size / 2;
1055
1056                 /* Limit pruning temp space to 20% of memory. Beyond this we discard
1057                  * the nodes and recompute them at the next move if necessary. */
1058                 u->max_pruned_size = u->max_tree_size / 5;
1059                 u->max_tree_size -= u->max_pruned_size;
1060         } else {
1061                 /* Reserve 5% memory in case the background free() are slower
1062                  * than the concurrent allocations. */
1063                 u->max_tree_size -= u->max_tree_size / 20;
1064         }
1065
1066         if (!u->prior)
1067                 u->prior = uct_prior_init(NULL, b, u);
1068
1069         if (!u->playout)
1070                 u->playout = playout_moggy_init(NULL, b, u->jdict);
1071         if (!u->playout->debug_level)
1072                 u->playout->debug_level = u->debug_level;
1073
1074         if (u->want_pat && !pat_setup)
1075                 patterns_init(&u->pat, NULL, false, true);
1076
1077         u->ownermap.map = malloc2(board_size2(b) * sizeof(u->ownermap.map[0]));
1078
1079         if (u->slave) {
1080                 if (!u->stats_hbits) u->stats_hbits = DEFAULT_STATS_HBITS;
1081                 if (!u->shared_nodes) u->shared_nodes = DEFAULT_SHARED_NODES;
1082                 assert(u->shared_levels * board_bits2(b) <= 8 * (int)sizeof(path_t));
1083         }
1084
1085         if (!u->dynkomi)
1086                 u->dynkomi = uct_dynkomi_init_linear(u, NULL, b);
1087
1088         /* Some things remain uninitialized for now - the opening tbook
1089          * is not loaded and the tree not set up. */
1090         /* This will be initialized in setup_state() at the first move
1091          * received/requested. This is because right now we are not aware
1092          * about any komi or handicap setup and such. */
1093
1094         return u;
1095 }
1096
1097 struct engine *
1098 engine_uct_init(char *arg, struct board *b)
1099 {
1100         struct uct *u = uct_state_init(arg, b);
1101         struct engine *e = calloc2(1, sizeof(struct engine));
1102         e->name = "UCT";
1103         e->printhook = uct_printhook_ownermap;
1104         e->notify_play = uct_notify_play;
1105         e->chat = uct_chat;
1106         e->undo = uct_undo;
1107         e->result = uct_result;
1108         e->genmove = uct_genmove;
1109         e->genmoves = uct_genmoves;
1110         e->evaluate = uct_evaluate;
1111         e->dead_group_list = uct_dead_group_list;
1112         e->done = uct_done;
1113         e->data = u;
1114         if (u->slave)
1115                 e->notify = uct_notify;
1116
1117         const char banner[] = "If you believe you have won but I am still playing, "
1118                 "please help me understand by capturing all dead stones. "
1119                 "Anyone can send me 'winrate' in private chat to get my assessment of the position.";
1120         if (!u->banner) u->banner = "";
1121         e->comment = malloc2(sizeof(banner) + strlen(u->banner) + 1);
1122         sprintf(e->comment, "%s %s", banner, u->banner);
1123
1124         return e;
1125 }