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