uct/uct.c

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