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 "chat.h"
  14 #include "move.h"
  15 #include "mq.h"
  16 #include "joseki/base.h"
  17 #include "playout.h"
  18 #include "playout/moggy.h"
  19 #include "playout/light.h"
  20 #include "tactics/util.h"
  21 #include "timeinfo.h"
  22 #include "uct/dynkomi.h"
  23 #include "uct/internal.h"
  24 #include "uct/plugins.h"
  25 #include "uct/prior.h"
  26 #include "uct/search.h"
  27 #include "uct/slave.h"
  28 #include "uct/tree.h"
  29 #include "uct/uct.h"
  30 #include "uct/walk.h"
  31
  32 struct uct_policy *policy_ucb1_init(struct uct *u, char *arg);
  33 struct uct_policy *policy_ucb1amaf_init(struct uct *u, char *arg, struct board *board);
  34 static void uct_pondering_start(struct uct *u, struct board *b0, struct tree *t, enum stone color);
  35
  36 /* Maximal simulation length. */
  37 #define MC_GAMELEN      MAX_GAMELEN
  38
  39
  40 static void
  41 setup_state(struct uct *u, struct board *b, enum stone color)
  42 {
  43         u->t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
  44                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, u->stats_hbits);
  45         if (u->force_seed)
  46                 fast_srandom(u->force_seed);
  47         if (UDEBUGL(3))
  48                 fprintf(stderr, "Fresh board with random seed %lu\n", fast_getseed());
  49         if (!u->no_tbook && b->moves == 0) {
  50                 if (color == S_BLACK) {
  51                         tree_load(u->t, b);
  52                 } else if (DEBUGL(0)) {
  53                         fprintf(stderr, "Warning: First move appears to be white\n");
  54                 }
  55         }
  56 }
  57
  58 static void
  59 reset_state(struct uct *u)
  60 {
  61         assert(u->t);
  62         tree_done(u->t); u->t = NULL;
  63 }
  64
  65 static void
  66 setup_dynkomi(struct uct *u, struct board *b, enum stone to_play)
  67 {
  68         if (u->t->use_extra_komi && !u->pondering && u->dynkomi->permove)
  69                 u->t->extra_komi = u->dynkomi->permove(u->dynkomi, b, u->t);
  70         else if (!u->t->use_extra_komi)
  71                 u->t->extra_komi = 0;
  72 }
  73
  74 void
  75 uct_prepare_move(struct uct *u, struct board *b, enum stone color)
  76 {
  77         if (u->t) {
  78                 /* Verify that we have sane state. */
  79                 assert(b->es == u);
  80                 assert(u->t && b->moves);
  81                 if (color != stone_other(u->t->root_color)) {
  82                         fprintf(stderr, "Fatal: Non-alternating play detected %d %d\n",
  83                                 color, u->t->root_color);
  84                         exit(1);
  85                 }
  86                 uct_htable_reset(u->t);
  87
  88         } else {
  89                 /* We need fresh state. */
  90                 b->es = u;
  91                 setup_state(u, b, color);
  92         }
  93
  94         u->ownermap.playouts = 0;
  95         memset(u->ownermap.map, 0, board_size2(b) * sizeof(u->ownermap.map[0]));
  96         u->played_own = u->played_all = 0;
  97 }
  98
  99 static void
 100 dead_group_list(struct uct *u, struct board *b, struct move_queue *mq)
 101 {
 102         enum gj_state gs_array[board_size2(b)];
 103         struct group_judgement gj = { .thres = GJ_THRES, .gs = gs_array };
 104         board_ownermap_judge_groups(b, &u->ownermap, &gj);
 105         groups_of_status(b, &gj, GS_DEAD, mq);
 106 }
 107
 108 bool
 109 uct_pass_is_safe(struct uct *u, struct board *b, enum stone color, bool pass_all_alive)
 110 {
 111         /* Make sure enough playouts are simulated to get a reasonable dead group list. */
 112         while (u->ownermap.playouts < GJ_MINGAMES)
 113                 uct_playout(u, b, color, u->t);
 114
 115         struct move_queue mq = { .moves = 0 };
 116         dead_group_list(u, b, &mq);
 117         if (pass_all_alive) {
 118                 for (unsigned int i = 0; i < mq.moves; i++) {
 119                         if (board_at(b, mq.move[i]) == stone_other(color)) {
 120                                 return false; // We need to remove opponent dead groups first.
 121                         }
 122                 }
 123                 mq.moves = 0; // our dead stones are alive when pass_all_alive is true
 124         }
 125         return pass_is_safe(b, color, &mq);
 126 }
 127
 128 static char *
 129 uct_printhook_ownermap(struct board *board, coord_t c, char *s, char *end)
 130 {
 131         struct uct *u = board->es;
 132         if (!u) {
 133                 strcat(s, ". ");
 134                 return s + 2;
 135         }
 136         const char chr[] = ":XO,"; // dame, black, white, unclear
 137         const char chm[] = ":xo,";
 138         char ch = chr[board_ownermap_judge_point(&u->ownermap, c, GJ_THRES)];
 139         if (ch == ',') { // less precise estimate then?
 140                 ch = chm[board_ownermap_judge_point(&u->ownermap, c, 0.67)];
 141         }
 142         s += snprintf(s, end - s, "%c ", ch);
 143         return s;
 144 }
 145
 146 static char *
 147 uct_notify_play(struct engine *e, struct board *b, struct move *m, char *enginearg)
 148 {
 149         struct uct *u = e->data;
 150         if (!u->t) {
 151                 /* No state, create one - this is probably game beginning
 152                  * and we need to load the opening tbook right now. */
 153                 uct_prepare_move(u, b, m->color);
 154                 assert(u->t);
 155         }
 156
 157         /* Stop pondering, required by tree_promote_at() */
 158         uct_pondering_stop(u);
 159         if (UDEBUGL(2) && u->slave)
 160                 tree_dump(u->t, u->dumpthres);
 161
 162         if (is_resign(m->coord)) {
 163                 /* Reset state. */
 164                 reset_state(u);
 165                 return NULL;
 166         }
 167
 168         /* Promote node of the appropriate move to the tree root. */
 169         assert(u->t->root);
 170         if (!tree_promote_at(u->t, b, m->coord)) {
 171                 if (UDEBUGL(3))
 172                         fprintf(stderr, "Warning: Cannot promote move node! Several play commands in row?\n");
 173                 reset_state(u);
 174                 return NULL;
 175         }
 176
 177         /* If we are a slave in a distributed engine, start pondering once
 178          * we know which move we actually played. See uct_genmove() about
 179          * the check for pass. */
 180         if (u->pondering_opt && u->slave && m->color == u->my_color && !is_pass(m->coord))
 181                 uct_pondering_start(u, b, u->t, stone_other(m->color));
 182
 183         return NULL;
 184 }
 185
 186 static char *
 187 uct_undo(struct engine *e, struct board *b)
 188 {
 189         struct uct *u = e->data;
 190
 191         if (!u->t) return NULL;
 192         uct_pondering_stop(u);
 193         reset_state(u);
 194         return NULL;
 195 }
 196
 197 static char *
 198 uct_result(struct engine *e, struct board *b)
 199 {
 200         struct uct *u = e->data;
 201         static char reply[1024];
 202
 203         if (!u->t)
 204                 return NULL;
 205         enum stone color = u->t->root_color;
 206         struct tree_node *n = u->t->root;
 207         snprintf(reply, 1024, "%s %s %d %.2f %.1f",
 208                  stone2str(color), coord2sstr(node_coord(n), b),
 209                  n->u.playouts, tree_node_get_value(u->t, -1, n->u.value),
 210                  u->t->use_extra_komi ? u->t->extra_komi : 0);
 211         return reply;
 212 }
 213
 214 static char *
 215 uct_chat(struct engine *e, struct board *b, bool opponent, char *from, char *cmd)
 216 {
 217         struct uct *u = e->data;
 218
 219         if (!u->t)
 220                 return generic_chat(b, opponent, from, cmd, S_NONE, pass, 0, 1, u->threads, 0.0, 0.0);
 221
 222         struct tree_node *n = u->t->root;
 223         double winrate = tree_node_get_value(u->t, -1, n->u.value);
 224         double extra_komi = u->t->use_extra_komi && abs(u->t->extra_komi) >= 0.5 ? u->t->extra_komi : 0;
 225
 226         return generic_chat(b, opponent, from, cmd, u->t->root_color, node_coord(n), n->u.playouts, 1,
 227                             u->threads, winrate, extra_komi);
 228 }
 229
 230 static void
 231 uct_dead_group_list(struct engine *e, struct board *b, struct move_queue *mq)
 232 {
 233         struct uct *u = e->data;
 234
 235         /* This means the game is probably over, no use pondering on. */
 236         uct_pondering_stop(u);
 237
 238         if (u->pass_all_alive)
 239                 return; // no dead groups
 240
 241         bool mock_state = false;
 242
 243         if (!u->t) {
 244                 /* No state, but we cannot just back out - we might
 245                  * have passed earlier, only assuming some stones are
 246                  * dead, and then re-connected, only to lose counting
 247                  * when all stones are assumed alive. */
 248                 uct_prepare_move(u, b, S_BLACK); assert(u->t);
 249                 mock_state = true;
 250         }
 251         /* Make sure the ownermap is well-seeded. */
 252         while (u->ownermap.playouts < GJ_MINGAMES)
 253                 uct_playout(u, b, S_BLACK, u->t);
 254         /* Show the ownermap: */
 255         if (DEBUGL(2))
 256                 board_print_custom(b, stderr, uct_printhook_ownermap);
 257
 258         dead_group_list(u, b, mq);
 259
 260         if (mock_state) {
 261                 /* Clean up the mock state in case we will receive
 262                  * a genmove; we could get a non-alternating-move
 263                  * error from uct_prepare_move() in that case otherwise. */
 264                 reset_state(u);
 265         }
 266 }
 267
 268 static void
 269 playout_policy_done(struct playout_policy *p)
 270 {
 271         if (p->done) p->done(p);
 272         if (p->data) free(p->data);
 273         free(p);
 274 }
 275
 276 static void
 277 uct_done(struct engine *e)
 278 {
 279         /* This is called on engine reset, especially when clear_board
 280          * is received and new game should begin. */
 281         struct uct *u = e->data;
 282         uct_pondering_stop(u);
 283         if (u->t) reset_state(u);
 284         free(u->ownermap.map);
 285
 286         free(u->policy);
 287         free(u->random_policy);
 288         playout_policy_done(u->playout);
 289         uct_prior_done(u->prior);
 290         joseki_done(u->jdict);
 291         pluginset_done(u->plugins);
 292 }
 293
 294
 295
 296 /* Run time-limited MCTS search on foreground. */
 297 static int
 298 uct_search(struct uct *u, struct board *b, struct time_info *ti, enum stone color, struct tree *t, bool print_progress)
 299 {
 300         struct uct_search_state s;
 301         uct_search_start(u, b, color, t, ti, &s);
 302         if (UDEBUGL(2) && s.base_playouts > 0)
 303                 fprintf(stderr, "<pre-simulated %d games>\n", s.base_playouts);
 304
 305         /* The search tree is ctx->t. This is currently == . It is important
 306          * to reference ctx->t directly since the
 307          * thread manager will swap the tree pointer asynchronously. */
 308
 309         /* Now, just periodically poll the search tree. */
 310         /* Note that in case of TD_GAMES, threads will not wait for
 311          * the uct_search_check_stop() signalization. */
 312         while (1) {
 313                 time_sleep(TREE_BUSYWAIT_INTERVAL);
 314                 /* TREE_BUSYWAIT_INTERVAL should never be less than desired time, or the
 315                  * time control is broken. But if it happens to be less, we still search
 316                  * at least 100ms otherwise the move is completely random. */
 317
 318                 int i = uct_search_games(&s);
 319                 /* Print notifications etc. */
 320                 uct_search_progress(u, b, color, t, ti, &s, i);
 321                 /* Check if we should stop the search. */
 322                 if (uct_search_check_stop(u, b, color, t, ti, &s, i))
 323                         break;
 324         }
 325
 326         struct uct_thread_ctx *ctx = uct_search_stop();
 327         if (UDEBUGL(2)) tree_dump(t, u->dumpthres);
 328         if (UDEBUGL(2))
 329                 fprintf(stderr, "(avg score %f/%d; dynkomi's %f/%d value %f/%d)\n",
 330                         t->avg_score.value, t->avg_score.playouts,
 331                         u->dynkomi->score.value, u->dynkomi->score.playouts,
 332                         u->dynkomi->value.value, u->dynkomi->value.playouts);
 333         if (print_progress)
 334                 uct_progress_status(u, t, color, ctx->games, NULL);
 335
 336         u->played_own += ctx->games;
 337         return ctx->games;
 338 }
 339
 340 /* Start pondering background with @color to play. */
 341 static void
 342 uct_pondering_start(struct uct *u, struct board *b0, struct tree *t, enum stone color)
 343 {
 344         if (UDEBUGL(1))
 345                 fprintf(stderr, "Starting to ponder with color %s\n", stone2str(stone_other(color)));
 346         u->pondering = true;
 347
 348         /* We need a local board copy to ponder upon. */
 349         struct board *b = malloc2(sizeof(*b)); board_copy(b, b0);
 350
 351         /* *b0 did not have the genmove'd move played yet. */
 352         struct move m = { node_coord(t->root), t->root_color };
 353         int res = board_play(b, &m);
 354         assert(res >= 0);
 355         setup_dynkomi(u, b, stone_other(m.color));
 356
 357         /* Start MCTS manager thread "headless". */
 358         static struct uct_search_state s;
 359         uct_search_start(u, b, color, t, NULL, &s);
 360 }
 361
 362 /* uct_search_stop() frontend for the pondering (non-genmove) mode, and
 363  * to stop the background search for a slave in the distributed engine. */
 364 void
 365 uct_pondering_stop(struct uct *u)
 366 {
 367         if (!thread_manager_running)
 368                 return;
 369
 370         /* Stop the thread manager. */
 371         struct uct_thread_ctx *ctx = uct_search_stop();
 372         if (UDEBUGL(1)) {
 373                 if (u->pondering) fprintf(stderr, "(pondering) ");
 374                 uct_progress_status(u, ctx->t, ctx->color, ctx->games, NULL);
 375         }
 376         if (u->pondering) {
 377                 free(ctx->b);
 378                 u->pondering = false;
 379         }
 380 }
 381
 382
 383 void
 384 uct_genmove_setup(struct uct *u, struct board *b, enum stone color)
 385 {
 386         if (b->superko_violation) {
 387                 fprintf(stderr, "!!! WARNING: SUPERKO VIOLATION OCCURED BEFORE THIS MOVE\n");
 388                 fprintf(stderr, "Maybe you play with situational instead of positional superko?\n");
 389                 fprintf(stderr, "I'm going to ignore the violation, but note that I may miss\n");
 390                 fprintf(stderr, "some moves valid under this ruleset because of this.\n");
 391                 b->superko_violation = false;
 392         }
 393
 394         uct_prepare_move(u, b, color);
 395
 396         assert(u->t);
 397         u->my_color = color;
 398
 399         /* How to decide whether to use dynkomi in this game? Since we use
 400          * pondering, it's not simple "who-to-play" matter. Decide based on
 401          * the last genmove issued. */
 402         u->t->use_extra_komi = !!(u->dynkomi_mask & color);
 403         setup_dynkomi(u, b, color);
 404
 405         if (b->rules == RULES_JAPANESE)
 406                 u->territory_scoring = true;
 407
 408         /* Make pessimistic assumption about komi for Japanese rules to
 409          * avoid losing by 0.5 when winning by 0.5 with Chinese rules.
 410          * The rules usually give the same winner if the integer part of komi
 411          * is odd so we adjust the komi only if it is even (for a board of
 412          * odd size). We are not trying  to get an exact evaluation for rare
 413          * cases of seki. For details see http://home.snafu.de/jasiek/parity.html */
 414         if (u->territory_scoring && (((int)floor(b->komi) + board_size(b)) & 1)) {
 415                 b->komi += (color == S_BLACK ? 1.0 : -1.0);
 416                 if (UDEBUGL(0))
 417                         fprintf(stderr, "Setting komi to %.1f assuming Japanese rules\n",
 418                                 b->komi);
 419         }
 420 }
 421
 422 static coord_t *
 423 uct_genmove(struct engine *e, struct board *b, struct time_info *ti, enum stone color, bool pass_all_alive)
 424 {
 425         double start_time = time_now();
 426         struct uct *u = e->data;
 427         u->pass_all_alive |= pass_all_alive;
 428         uct_pondering_stop(u);
 429         uct_genmove_setup(u, b, color);
 430
 431         /* Start the Monte Carlo Tree Search! */
 432         int base_playouts = u->t->root->u.playouts;
 433         int played_games = uct_search(u, b, ti, color, u->t, false);
 434
 435         coord_t best_coord;
 436         struct tree_node *best;
 437         best = uct_search_result(u, b, color, u->pass_all_alive, played_games, base_playouts, &best_coord);
 438
 439         if (UDEBUGL(2)) {
 440                 double time = time_now() - start_time + 0.000001; /* avoid divide by zero */
 441                 fprintf(stderr, "genmove in %0.2fs (%d games/s, %d games/s/thread)\n",
 442                         time, (int)(played_games/time), (int)(played_games/time/u->threads));
 443         }
 444
 445         uct_progress_status(u, u->t, color, played_games, &best_coord);
 446
 447         if (!best) {
 448                 /* Pass or resign. */
 449                 reset_state(u);
 450                 return coord_copy(best_coord);
 451         }
 452         tree_promote_node(u->t, &best);
 453
 454         /* After a pass, pondering is harmful for two reasons:
 455          * (i) We might keep pondering even when the game is over.
 456          * Of course this is the case for opponent resign as well.
 457          * (ii) More importantly, the ownermap will get skewed since
 458          * the UCT will start cutting off any playouts. */
 459         if (u->pondering_opt && !is_pass(node_coord(best))) {
 460                 uct_pondering_start(u, b, u->t, stone_other(color));
 461         }
 462         return coord_copy(best_coord);
 463 }
 464
 465
 466 bool
 467 uct_gentbook(struct engine *e, struct board *b, struct time_info *ti, enum stone color)
 468 {
 469         struct uct *u = e->data;
 470         if (!u->t) uct_prepare_move(u, b, color);
 471         assert(u->t);
 472
 473         if (ti->dim == TD_GAMES) {
 474                 /* Don't count in games that already went into the tbook. */
 475                 ti->len.games += u->t->root->u.playouts;
 476         }
 477         uct_search(u, b, ti, color, u->t, true);
 478
 479         assert(ti->dim == TD_GAMES);
 480         tree_save(u->t, b, ti->len.games / 100);
 481
 482         return true;
 483 }
 484
 485 void
 486 uct_dumptbook(struct engine *e, struct board *b, enum stone color)
 487 {
 488         struct uct *u = e->data;
 489         struct tree *t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
 490                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, 0);
 491         tree_load(t, b);
 492         tree_dump(t, 0);
 493         tree_done(t);
 494 }
 495
 496
 497 floating_t
 498 uct_evaluate_one(struct engine *e, struct board *b, struct time_info *ti, coord_t c, enum stone color)
 499 {
 500         struct uct *u = e->data;
 501
 502         struct board b2;
 503         board_copy(&b2, b);
 504         struct move m = { c, color };
 505         int res = board_play(&b2, &m);
 506         if (res < 0)
 507                 return NAN;
 508         color = stone_other(color);
 509
 510         if (u->t) reset_state(u);
 511         uct_prepare_move(u, &b2, color);
 512         assert(u->t);
 513
 514         floating_t bestval;
 515         uct_search(u, &b2, ti, color, u->t, true);
 516         struct tree_node *best = u->policy->choose(u->policy, u->t->root, &b2, color, resign);
 517         if (!best) {
 518                 bestval = NAN; // the opponent has no reply!
 519         } else {
 520                 bestval = tree_node_get_value(u->t, 1, best->u.value);
 521         }
 522
 523         reset_state(u); // clean our junk
 524
 525         return isnan(bestval) ? NAN : 1.0f - bestval;
 526 }
 527
 528 void
 529 uct_evaluate(struct engine *e, struct board *b, struct time_info *ti, floating_t *vals, enum stone color)
 530 {
 531         for (int i = 0; i < b->flen; i++) {
 532                 if (is_pass(b->f[i]))
 533                         vals[i] = NAN;
 534                 else
 535                         vals[i] = uct_evaluate_one(e, b, ti, b->f[i], color);
 536         }
 537 }
 538
 539
 540 struct uct *
 541 uct_state_init(char *arg, struct board *b)
 542 {
 543         struct uct *u = calloc2(1, sizeof(struct uct));
 544         bool pat_setup = false;
 545
 546         u->debug_level = debug_level;
 547         u->reportfreq = 10000;
 548         u->gamelen = MC_GAMELEN;
 549         u->resign_threshold = 0.2;
 550         u->sure_win_threshold = 0.9;
 551         u->mercymin = 0;
 552         u->significant_threshold = 50;
 553         u->expand_p = 8;
 554         u->dumpthres = 1000;
 555         u->playout_amaf = true;
 556         u->amaf_prior = false;
 557         u->max_tree_size = 1408ULL * 1048576;
 558         u->fast_alloc = true;
 559         u->pruning_threshold = 0;
 560
 561         u->threads = 1;
 562         u->thread_model = TM_TREEVL;
 563         u->virtual_loss = 1;
 564
 565         u->fuseki_end = 20; // max time at 361*20% = 72 moves (our 36th move, still 99 to play)
 566         u->yose_start = 40; // (100-40-25)*361/100/2 = 63 moves still to play by us then
 567         u->bestr_ratio = 0.02;
 568         // 2.5 is clearly too much, but seems to compensate well for overly stern time allocations.
 569         // TODO: Further tuning and experiments with better time allocation schemes.
 570         u->best2_ratio = 2.5;
 571         u->max_maintime_ratio = 3.0;
 572
 573         u->val_scale = 0; u->val_points = 40;
 574         u->dynkomi_interval = 1000;
 575         u->dynkomi_mask = S_BLACK | S_WHITE;
 576
 577         u->tenuki_d = 4;
 578         u->local_tree_aging = 80;
 579         u->local_tree_depth_decay = 1.5;
 580         u->local_tree_eval = LTE_ROOT;
 581         u->local_tree_neival = true;
 582
 583         u->max_slaves = -1;
 584         u->slave_index = -1;
 585         u->stats_delay = 0.01; // 10 ms
 586
 587         u->plugins = pluginset_init(b);
 588
 589         u->jdict = joseki_load(b->size);
 590
 591         if (arg) {
 592                 char *optspec, *next = arg;
 593                 while (*next) {
 594                         optspec = next;
 595                         next += strcspn(next, ",");
 596                         if (*next) { *next++ = 0; } else { *next = 0; }
 597
 598                         char *optname = optspec;
 599                         char *optval = strchr(optspec, '=');
 600                         if (optval) *optval++ = 0;
 601
 602                         /** Basic options */
 603
 604                         if (!strcasecmp(optname, "debug")) {
 605                                 if (optval)
 606                                         u->debug_level = atoi(optval);
 607                                 else
 608                                         u->debug_level++;
 609                         } else if (!strcasecmp(optname, "reporting") && optval) {
 610                                 /* The format of output for detailed progress
 611                                  * information (such as current best move and
 612                                  * its value, etc.). */
 613                                 if (!strcasecmp(optval, "text")) {
 614                                         /* Plaintext traditional output. */
 615                                         u->reporting = UR_TEXT;
 616                                 } else if (!strcasecmp(optval, "json")) {
 617                                         /* JSON output. Implies debug=0. */
 618                                         u->reporting = UR_JSON;
 619                                         u->debug_level = 0;
 620                                 } else if (!strcasecmp(optval, "jsonbig")) {
 621                                         /* JSON output, but much more detailed.
 622                                          * Implies debug=0. */
 623                                         u->reporting = UR_JSON_BIG;
 624                                         u->debug_level = 0;
 625                                 } else {
 626                                         fprintf(stderr, "UCT: Invalid reporting format %s\n", optval);
 627                                         exit(1);
 628                                 }
 629                         } else if (!strcasecmp(optname, "reportfreq") && optval) {
 630                                 /* The progress information line will be shown
 631                                  * every <reportfreq> simulations. */
 632                                 u->reportfreq = atoi(optval);
 633                         } else if (!strcasecmp(optname, "dumpthres") && optval) {
 634                                 /* When dumping the UCT tree on output, include
 635                                  * nodes with at least this many playouts.
 636                                  * (This value is re-scaled "intelligently"
 637                                  * in case of very large trees.) */
 638                                 u->dumpthres = atoi(optval);
 639                         } else if (!strcasecmp(optname, "resign_threshold") && optval) {
 640                                 /* Resign when this ratio of games is lost
 641                                  * after GJ_MINGAMES sample is taken. */
 642                                 u->resign_threshold = atof(optval);
 643                         } else if (!strcasecmp(optname, "sure_win_threshold") && optval) {
 644                                 /* Stop reading when this ratio of games is won
 645                                  * after PLAYOUT_EARLY_BREAK_MIN sample is
 646                                  * taken. (Prevents stupid time losses,
 647                                  * friendly to human opponents.) */
 648                                 u->sure_win_threshold = atof(optval);
 649                         } else if (!strcasecmp(optname, "force_seed") && optval) {
 650                                 /* Set RNG seed at the tree setup. */
 651                                 u->force_seed = atoi(optval);
 652                         } else if (!strcasecmp(optname, "no_tbook")) {
 653                                 /* Disable UCT opening tbook. */
 654                                 u->no_tbook = true;
 655                         } else if (!strcasecmp(optname, "pass_all_alive")) {
 656                                 /* Whether to consider passing only after all
 657                                  * dead groups were removed from the board;
 658                                  * this is like all genmoves are in fact
 659                                  * kgs-genmove_cleanup. */
 660                                 u->pass_all_alive = !optval || atoi(optval);
 661                         } else if (!strcasecmp(optname, "territory_scoring")) {
 662                                 /* Use territory scoring (default is area scoring).
 663                                  * An explicit kgs-rules command overrides this. */
 664                                 u->territory_scoring = !optval || atoi(optval);
 665                         } else if (!strcasecmp(optname, "stones_only")) {
 666                                 /* Do not count eyes. Nice to teach go to kids.
 667                                  * http://strasbourg.jeudego.org/regle_strasbourgeoise.htm */
 668                                 b->rules = RULES_STONES_ONLY;
 669                                 u->pass_all_alive = true;
 670                         } else if (!strcasecmp(optname, "banner") && optval) {
 671                                 /* Additional banner string. This must come as the
 672                                  * last engine parameter. */
 673                                 if (*next) *--next = ',';
 674                                 u->banner = strdup(optval);
 675                                 break;
 676                         } else if (!strcasecmp(optname, "plugin") && optval) {
 677                                 /* Load an external plugin; filename goes before the colon,
 678                                  * extra arguments after the colon. */
 679                                 char *pluginarg = strchr(optval, ':');
 680                                 if (pluginarg)
 681                                         *pluginarg++ = 0;
 682                                 plugin_load(u->plugins, optval, pluginarg);
 683
 684                         /** UCT behavior and policies */
 685
 686                         } else if ((!strcasecmp(optname, "policy")
 687                                 /* Node selection policy. ucb1amaf is the
 688                                  * default policy implementing RAVE, while
 689                                  * ucb1 is the simple exploration/exploitation
 690                                  * policy. Policies can take further extra
 691                                  * options. */
 692                                     || !strcasecmp(optname, "random_policy")) && optval) {
 693                                 /* A policy to be used randomly with small
 694                                  * chance instead of the default policy. */
 695                                 char *policyarg = strchr(optval, ':');
 696                                 struct uct_policy **p = !strcasecmp(optname, "policy") ? &u->policy : &u->random_policy;
 697                                 if (policyarg)
 698                                         *policyarg++ = 0;
 699                                 if (!strcasecmp(optval, "ucb1")) {
 700                                         *p = policy_ucb1_init(u, policyarg);
 701                                 } else if (!strcasecmp(optval, "ucb1amaf")) {
 702                                         *p = policy_ucb1amaf_init(u, policyarg, b);
 703                                 } else {
 704                                         fprintf(stderr, "UCT: Invalid tree policy %s\n", optval);
 705                                         exit(1);
 706                                 }
 707                         } else if (!strcasecmp(optname, "playout") && optval) {
 708                                 /* Random simulation (playout) policy.
 709                                  * moggy is the default policy with large
 710                                  * amount of domain-specific knowledge and
 711                                  * heuristics. light is a simple uniformly
 712                                  * random move selection policy. */
 713                                 char *playoutarg = strchr(optval, ':');
 714                                 if (playoutarg)
 715                                         *playoutarg++ = 0;
 716                                 if (!strcasecmp(optval, "moggy")) {
 717                                         u->playout = playout_moggy_init(playoutarg, b, u->jdict);
 718                                 } else if (!strcasecmp(optval, "light")) {
 719                                         u->playout = playout_light_init(playoutarg, b);
 720                                 } else {
 721                                         fprintf(stderr, "UCT: Invalid playout policy %s\n", optval);
 722                                         exit(1);
 723                                 }
 724                         } else if (!strcasecmp(optname, "prior") && optval) {
 725                                 /* Node priors policy. When expanding a node,
 726                                  * it will seed node values heuristically
 727                                  * (most importantly, based on playout policy
 728                                  * opinion, but also with regard to other
 729                                  * things). See uct/prior.c for details.
 730                                  * Use prior=eqex=0 to disable priors. */
 731                                 u->prior = uct_prior_init(optval, b, u);
 732                         } else if (!strcasecmp(optname, "mercy") && optval) {
 733                                 /* Minimal difference of black/white captures
 734                                  * to stop playout - "Mercy Rule". Speeds up
 735                                  * hopeless playouts at the expense of some
 736                                  * accuracy. */
 737                                 u->mercymin = atoi(optval);
 738                         } else if (!strcasecmp(optname, "gamelen") && optval) {
 739                                 /* Maximum length of single simulation
 740                                  * in moves. */
 741                                 u->gamelen = atoi(optval);
 742                         } else if (!strcasecmp(optname, "expand_p") && optval) {
 743                                 /* Expand UCT nodes after it has been
 744                                  * visited this many times. */
 745                                 u->expand_p = atoi(optval);
 746                         } else if (!strcasecmp(optname, "random_policy_chance") && optval) {
 747                                 /* If specified (N), with probability 1/N, random_policy policy
 748                                  * descend is used instead of main policy descend; useful
 749                                  * if specified policy (e.g. UCB1AMAF) can make unduly biased
 750                                  * choices sometimes, you can fall back to e.g.
 751                                  * random_policy=UCB1. */
 752                                 u->random_policy_chance = atoi(optval);
 753
 754                         /** General AMAF behavior */
 755                         /* (Only relevant if the policy supports AMAF.
 756                          * More variables can be tuned as policy
 757                          * parameters.) */
 758
 759                         } else if (!strcasecmp(optname, "playout_amaf")) {
 760                                 /* Whether to include random playout moves in
 761                                  * AMAF as well. (Otherwise, only tree moves
 762                                  * are included in AMAF. Of course makes sense
 763                                  * only in connection with an AMAF policy.) */
 764                                 /* with-without: 55.5% (+-4.1) */
 765                                 if (optval && *optval == '0')
 766                                         u->playout_amaf = false;
 767                                 else
 768                                         u->playout_amaf = true;
 769                         } else if (!strcasecmp(optname, "playout_amaf_cutoff") && optval) {
 770                                 /* Keep only first N% of playout stage AMAF
 771                                  * information. */
 772                                 u->playout_amaf_cutoff = atoi(optval);
 773                         } else if (!strcasecmp(optname, "amaf_prior") && optval) {
 774                                 /* In node policy, consider prior values
 775                                  * part of the real result term or part
 776                                  * of the AMAF term? */
 777                                 u->amaf_prior = atoi(optval);
 778
 779                         /** Performance and memory management */
 780
 781                         } else if (!strcasecmp(optname, "threads") && optval) {
 782                                 /* By default, Pachi will run with only single
 783                                  * tree search thread! */
 784                                 u->threads = atoi(optval);
 785                         } else if (!strcasecmp(optname, "thread_model") && optval) {
 786                                 if (!strcasecmp(optval, "tree")) {
 787                                         /* Tree parallelization - all threads
 788                                          * grind on the same tree. */
 789                                         u->thread_model = TM_TREE;
 790                                         u->virtual_loss = 0;
 791                                 } else if (!strcasecmp(optval, "treevl")) {
 792                                         /* Tree parallelization, but also
 793                                          * with virtual losses - this discou-
 794                                          * rages most threads choosing the
 795                                          * same tree branches to read. */
 796                                         u->thread_model = TM_TREEVL;
 797                                 } else {
 798                                         fprintf(stderr, "UCT: Invalid thread model %s\n", optval);
 799                                         exit(1);
 800                                 }
 801                         } else if (!strcasecmp(optname, "virtual_loss")) {
 802                                 /* Number of virtual losses added before evaluating a node. */
 803                                 u->virtual_loss = !optval || atoi(optval);
 804                         } else if (!strcasecmp(optname, "pondering")) {
 805                                 /* Keep searching even during opponent's turn. */
 806                                 u->pondering_opt = !optval || atoi(optval);
 807                         } else if (!strcasecmp(optname, "max_tree_size") && optval) {
 808                                 /* Maximum amount of memory [MiB] consumed by the move tree.
 809                                  * For fast_alloc it includes the temp tree used for pruning.
 810                                  * Default is 3072 (3 GiB). */
 811                                 u->max_tree_size = atol(optval) * 1048576;
 812                         } else if (!strcasecmp(optname, "fast_alloc")) {
 813                                 u->fast_alloc = !optval || atoi(optval);
 814                         } else if (!strcasecmp(optname, "pruning_threshold") && optval) {
 815                                 /* Force pruning at beginning of a move if the tree consumes
 816                                  * more than this [MiB]. Default is 10% of max_tree_size.
 817                                  * Increase to reduce pruning time overhead if memory is plentiful.
 818                                  * This option is meaningful only for fast_alloc. */
 819                                 u->pruning_threshold = atol(optval) * 1048576;
 820
 821                         /** Time control */
 822
 823                         } else if (!strcasecmp(optname, "best2_ratio") && optval) {
 824                                 /* If set, prolong simulating while
 825                                  * first_best/second_best playouts ratio
 826                                  * is less than best2_ratio. */
 827                                 u->best2_ratio = atof(optval);
 828                         } else if (!strcasecmp(optname, "bestr_ratio") && optval) {
 829                                 /* If set, prolong simulating while
 830                                  * best,best_best_child values delta
 831                                  * is more than bestr_ratio. */
 832                                 u->bestr_ratio = atof(optval);
 833                         } else if (!strcasecmp(optname, "max_maintime_ratio") && optval) {
 834                                 /* If set and while not in byoyomi, prolong simulating no more than
 835                                  * max_maintime_ratio times the normal desired thinking time. */
 836                                 u->max_maintime_ratio = atof(optval);
 837                         } else if (!strcasecmp(optname, "fuseki_end") && optval) {
 838                                 /* At the very beginning it's not worth thinking
 839                                  * too long because the playout evaluations are
 840                                  * very noisy. So gradually increase the thinking
 841                                  * time up to maximum when fuseki_end percent
 842                                  * of the board has been played.
 843                                  * This only applies if we are not in byoyomi. */
 844                                 u->fuseki_end = atoi(optval);
 845                         } else if (!strcasecmp(optname, "yose_start") && optval) {
 846                                 /* When yose_start percent of the board has been
 847                                  * played, or if we are in byoyomi, stop spending
 848                                  * more time and spread the remaining time
 849                                  * uniformly.
 850                                  * Between fuseki_end and yose_start, we spend
 851                                  * a constant proportion of the remaining time
 852                                  * on each move. (yose_start should actually
 853                                  * be much earlier than when real yose start,
 854                                  * but "yose" is a good short name to convey
 855                                  * the idea.) */
 856                                 u->yose_start = atoi(optval);
 857
 858                         /** Dynamic komi */
 859
 860                         } else if (!strcasecmp(optname, "dynkomi") && optval) {
 861                                 /* Dynamic komi approach; there are multiple
 862                                  * ways to adjust komi dynamically throughout
 863                                  * play. We currently support two: */
 864                                 char *dynkomiarg = strchr(optval, ':');
 865                                 if (dynkomiarg)
 866                                         *dynkomiarg++ = 0;
 867                                 if (!strcasecmp(optval, "none")) {
 868                                         u->dynkomi = uct_dynkomi_init_none(u, dynkomiarg, b);
 869                                 } else if (!strcasecmp(optval, "linear")) {
 870                                         /* You should set dynkomi_mask=1 or a very low
 871                                          * handicap_value for white. */
 872                                         u->dynkomi = uct_dynkomi_init_linear(u, dynkomiarg, b);
 873                                 } else if (!strcasecmp(optval, "adaptive")) {
 874                                         /* There are many more knobs to
 875                                          * crank - see uct/dynkomi.c. */
 876                                         u->dynkomi = uct_dynkomi_init_adaptive(u, dynkomiarg, b);
 877                                 } else {
 878                                         fprintf(stderr, "UCT: Invalid dynkomi mode %s\n", optval);
 879                                         exit(1);
 880                                 }
 881                         } else if (!strcasecmp(optname, "dynkomi_mask") && optval) {
 882                                 /* Bitmask of colors the player must be
 883                                  * for dynkomi be applied; the default dynkomi_mask=3 allows
 884                                  * dynkomi 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")) {
 914                                 /* Whether to bias exploration by local tree values. */
 915                                 u->local_tree = !optval || atoi(optval);
 916                         } else if (!strcasecmp(optname, "tenuki_d") && optval) {
 917                                 /* Tenuki distance at which to break the local tree. */
 918                                 u->tenuki_d = atoi(optval);
 919                                 if (u->tenuki_d > TREE_NODE_D_MAX + 1) {
 920                                         fprintf(stderr, "uct: tenuki_d must not be larger than TREE_NODE_D_MAX+1 %d\n", TREE_NODE_D_MAX + 1);
 921                                         exit(1);
 922                                 }
 923                         } else if (!strcasecmp(optname, "local_tree_aging") && optval) {
 924                                 /* How much to reduce local tree values between moves. */
 925                                 u->local_tree_aging = atof(optval);
 926                         } else if (!strcasecmp(optname, "local_tree_depth_decay") && optval) {
 927                                 /* With value x>0, during the descent the node
 928                                  * contributes 1/x^depth playouts in
 929                                  * the local tree. I.e., with x>1, nodes more
 930                                  * distant from local situation contribute more
 931                                  * than nodes near the root. */
 932                                 u->local_tree_depth_decay = atof(optval);
 933                         } else if (!strcasecmp(optname, "local_tree_allseq")) {
 934                                 /* If disabled, only complete sequences are stored
 935                                  * in the local tree. If this is on, also
 936                                  * subsequences starting at each move are stored. */
 937                                 u->local_tree_allseq = !optval || atoi(optval);
 938                         } else if (!strcasecmp(optname, "local_tree_neival")) {
 939                                 /* If disabled, local node value is not
 940                                  * computed just based on terminal status
 941                                  * of the coordinate, but also its neighbors. */
 942                                 u->local_tree_neival = !optval || atoi(optval);
 943                         } else if (!strcasecmp(optname, "local_tree_eval")) {
 944                                 /* How is the value inserted in the local tree
 945                                  * determined. */
 946                                 if (!strcasecmp(optval, "root"))
 947                                         /* All moves within a tree branch are
 948                                          * considered wrt. their merit
 949                                          * reaching tachtical goal of making
 950                                          * the first move in the branch
 951                                          * survive. */
 952                                         u->local_tree_eval = LTE_ROOT;
 953                                 else if (!strcasecmp(optval, "each"))
 954                                         /* Each move is considered wrt.
 955                                          * its own survival. */
 956                                         u->local_tree_eval = LTE_EACH;
 957                                 else if (!strcasecmp(optval, "total"))
 958                                         /* The tactical goal is the survival
 959                                          * of all the moves of my color and
 960                                          * non-survival of all the opponent
 961                                          * moves. Local values (and their
 962                                          * inverses) are averaged. */
 963                                         u->local_tree_eval = LTE_TOTAL;
 964                                 else {
 965                                         fprintf(stderr, "uct: unknown local_tree_eval %s\n", optval);
 966                                         exit(1);
 967                                 }
 968                         } else if (!strcasecmp(optname, "local_tree_rootchoose")) {
 969                                 /* If disabled, only moves within the local
 970                                  * tree branch are considered; the values
 971                                  * of the branch roots (i.e. root children)
 972                                  * are ignored. This may make sense together
 973                                  * with eval!=each, we consider only moves
 974                                  * that influence the goal, not the "rating"
 975                                  * of the goal itself. (The real solution
 976                                  * will be probably using criticality to pick
 977                                  * local tree branches.) */
 978                                 u->local_tree_rootchoose = !optval || atoi(optval);
 979
 980                         /** Other heuristics */
 981                         } else if (!strcasecmp(optname, "patterns")) {
 982                                 /* Load pattern database. Various modules
 983                                  * (priors, policies etc.) may make use
 984                                  * of this database. They will request
 985                                  * it automatically in that case, but you
 986                                  * can use this option to tweak the pattern
 987                                  * parameters. */
 988                                 patterns_init(&u->pat, optval, false, true);
 989                                 u->want_pat = pat_setup = true;
 990                         } else if (!strcasecmp(optname, "significant_threshold") && optval) {
 991                                 /* Some heuristics (XXX: none in mainline) rely
 992                                  * on the knowledge of the last "significant"
 993                                  * node in the descent. Such a node is
 994                                  * considered reasonably trustworthy to carry
 995                                  * some meaningful information in the values
 996                                  * of the node and its children. */
 997                                 u->significant_threshold = atoi(optval);
 998
 999                         /** Distributed engine slaves setup */
1000
1001                         } else if (!strcasecmp(optname, "slave")) {
1002                                 /* Act as slave for the distributed engine. */
1003                                 u->slave = !optval || atoi(optval);
1004                         } else if (!strcasecmp(optname, "slave_index") && optval) {
1005                                 /* Optional index if per-slave behavior is desired.
1006                                  * Must be given as index/max */
1007                                 u->slave_index = atoi(optval);
1008                                 char *p = strchr(optval, '/');
1009                                 if (p) u->max_slaves = atoi(++p);
1010                         } else if (!strcasecmp(optname, "shared_nodes") && optval) {
1011                                 /* Share at most shared_nodes between master and slave at each genmoves.
1012                                  * Must use the same value in master and slaves. */
1013                                 u->shared_nodes = atoi(optval);
1014                         } else if (!strcasecmp(optname, "shared_levels") && optval) {
1015                                 /* Share only nodes of level <= shared_levels. */
1016                                 u->shared_levels = atoi(optval);
1017                         } else if (!strcasecmp(optname, "stats_hbits") && optval) {
1018                                 /* Set hash table size to 2^stats_hbits for the shared stats. */
1019                                 u->stats_hbits = atoi(optval);
1020                         } else if (!strcasecmp(optname, "stats_delay") && optval) {
1021                                 /* How long to wait in slave for initial stats to build up before
1022                                  * replying to the genmoves command (in ms) */
1023                                 u->stats_delay = 0.001 * atof(optval);
1024
1025                         } else {
1026                                 fprintf(stderr, "uct: Invalid engine argument %s or missing value\n", optname);
1027                                 exit(1);
1028                         }
1029                 }
1030         }
1031
1032         if (!u->policy)
1033                 u->policy = policy_ucb1amaf_init(u, NULL, b);
1034
1035         if (!!u->random_policy_chance ^ !!u->random_policy) {
1036                 fprintf(stderr, "uct: Only one of random_policy and random_policy_chance is set\n");
1037                 exit(1);
1038         }
1039
1040         if (!u->local_tree) {
1041                 /* No ltree aging. */
1042                 u->local_tree_aging = 1.0f;
1043         }
1044
1045         if (u->fast_alloc) {
1046                 if (u->pruning_threshold < u->max_tree_size / 10)
1047                         u->pruning_threshold = u->max_tree_size / 10;
1048                 if (u->pruning_threshold > u->max_tree_size / 2)
1049                         u->pruning_threshold = u->max_tree_size / 2;
1050
1051                 /* Limit pruning temp space to 20% of memory. Beyond this we discard
1052                  * the nodes and recompute them at the next move if necessary. */
1053                 u->max_pruned_size = u->max_tree_size / 5;
1054                 u->max_tree_size -= u->max_pruned_size;
1055         } else {
1056                 /* Reserve 5% memory in case the background free() are slower
1057                  * than the concurrent allocations. */
1058                 u->max_tree_size -= u->max_tree_size / 20;
1059         }
1060
1061         if (!u->prior)
1062                 u->prior = uct_prior_init(NULL, b, u);
1063
1064         if (!u->playout)
1065                 u->playout = playout_moggy_init(NULL, b, u->jdict);
1066         if (!u->playout->debug_level)
1067                 u->playout->debug_level = u->debug_level;
1068
1069         if (u->want_pat && !pat_setup)
1070                 patterns_init(&u->pat, NULL, false, true);
1071
1072         u->ownermap.map = malloc2(board_size2(b) * sizeof(u->ownermap.map[0]));
1073
1074         if (u->slave) {
1075                 if (!u->stats_hbits) u->stats_hbits = DEFAULT_STATS_HBITS;
1076                 if (!u->shared_nodes) u->shared_nodes = DEFAULT_SHARED_NODES;
1077                 assert(u->shared_levels * board_bits2(b) <= 8 * (int)sizeof(path_t));
1078         }
1079
1080         if (!u->dynkomi)
1081                 u->dynkomi = uct_dynkomi_init_linear(u, NULL, b);
1082
1083         /* Some things remain uninitialized for now - the opening tbook
1084          * is not loaded and the tree not set up. */
1085         /* This will be initialized in setup_state() at the first move
1086          * received/requested. This is because right now we are not aware
1087          * about any komi or handicap setup and such. */
1088
1089         return u;
1090 }
1091
1092 struct engine *
1093 engine_uct_init(char *arg, struct board *b)
1094 {
1095         struct uct *u = uct_state_init(arg, b);
1096         struct engine *e = calloc2(1, sizeof(struct engine));
1097         e->name = "UCT";
1098         e->printhook = uct_printhook_ownermap;
1099         e->notify_play = uct_notify_play;
1100         e->chat = uct_chat;
1101         e->undo = uct_undo;
1102         e->result = uct_result;
1103         e->genmove = uct_genmove;
1104         e->genmoves = uct_genmoves;
1105         e->evaluate = uct_evaluate;
1106         e->dead_group_list = uct_dead_group_list;
1107         e->done = uct_done;
1108         e->data = u;
1109         if (u->slave)
1110                 e->notify = uct_notify;
1111
1112         const char banner[] = "If you believe you have won but I am still playing, "
1113                 "please help me understand by capturing all dead stones. "
1114                 "Anyone can send me 'winrate' in private chat to get my assessment of the position.";
1115         if (!u->banner) u->banner = "";
1116         e->comment = malloc2(sizeof(banner) + strlen(u->banner) + 1);
1117         sprintf(e->comment, "%s %s", banner, u->banner);
1118
1119         return e;
1120 }