uct/uct.c

   1 #include <assert.h>
   2 #include <pthread.h>
   3 #include <signal.h>
   4 #include <stdio.h>
   5 #include <stdlib.h>
   6 #include <string.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 "playout.h"
  16 #include "playout/elo.h"
  17 #include "playout/moggy.h"
  18 #include "playout/light.h"
  19 #include "random.h"
  20 #include "tactics.h"
  21 #include "uct/internal.h"
  22 #include "uct/prior.h"
  23 #include "uct/tree.h"
  24 #include "uct/uct.h"
  25 #include "uct/walk.h"
  26
  27 struct uct_policy *policy_ucb1_init(struct uct *u, char *arg);
  28 struct uct_policy *policy_ucb1amaf_init(struct uct *u, char *arg);
  29 static void uct_pondering_finish(struct uct *u);
  30
  31
  32 #define MC_GAMES        80000
  33 #define MC_GAMELEN      MAX_GAMELEN
  34
  35 /* How big proportion of ownermap counts must be of one color to consider
  36  * the point sure. */
  37 #define GJ_THRES        0.8
  38 /* How many games to consider at minimum before judging groups. */
  39 #define GJ_MINGAMES     500
  40
  41
  42 static void
  43 setup_state(struct uct *u, struct board *b, enum stone color)
  44 {
  45         u->t = tree_init(b, color);
  46         if (u->force_seed)
  47                 fast_srandom(u->force_seed);
  48         if (UDEBUGL(0))
  49                 fprintf(stderr, "Fresh board with random seed %lu\n", fast_getseed());
  50         //board_print(b, stderr);
  51         if (!u->no_book && b->moves == 0) {
  52                 assert(color == S_BLACK);
  53                 tree_load(u->t, b);
  54         }
  55 }
  56
  57 static void
  58 reset_state(struct uct *u)
  59 {
  60         assert(u->t);
  61         tree_done(u->t); u->t = NULL;
  62 }
  63
  64 static void
  65 prepare_move(struct engine *e, struct board *b, enum stone color)
  66 {
  67         struct uct *u = e->data;
  68
  69         if (u->t) {
  70                 /* Verify that we have sane state. */
  71                 assert(b->es == u);
  72                 assert(u->t && b->moves);
  73                 if (color != stone_other(u->t->root_color)) {
  74                         fprintf(stderr, "Fatal: Non-alternating play detected %d %d\n",
  75                                 color, u->t->root_color);
  76                         exit(1);
  77                 }
  78
  79         } else {
  80                 /* We need fresh state. */
  81                 b->es = u;
  82                 setup_state(u, b, color);
  83         }
  84
  85         if (u->dynkomi && u->dynkomi > b->moves && (color & u->dynkomi_mask))
  86                 u->t->extra_komi = uct_get_extra_komi(u, b);
  87
  88         u->ownermap.playouts = 0;
  89         memset(u->ownermap.map, 0, board_size2(b) * sizeof(u->ownermap.map[0]));
  90 }
  91
  92 static void
  93 dead_group_list(struct uct *u, struct board *b, struct move_queue *mq)
  94 {
  95         struct group_judgement gj;
  96         gj.thres = GJ_THRES;
  97         gj.gs = alloca(board_size2(b) * sizeof(gj.gs[0]));
  98         board_ownermap_judge_group(b, &u->ownermap, &gj);
  99         groups_of_status(b, &gj, GS_DEAD, mq);
 100 }
 101
 102 bool
 103 uct_pass_is_safe(struct uct *u, struct board *b, enum stone color, bool pass_all_alive)
 104 {
 105         if (u->ownermap.playouts < GJ_MINGAMES)
 106                 return false;
 107
 108         struct move_queue mq = { .moves = 0 };
 109         if (!pass_all_alive)
 110                 dead_group_list(u, b, &mq);
 111         return pass_is_safe(b, color, &mq);
 112 }
 113
 114
 115 static void
 116 uct_printhook_ownermap(struct board *board, coord_t c, FILE *f)
 117 {
 118         struct uct *u = board->es;
 119         assert(u);
 120         const char chr[] = ":XO,"; // dame, black, white, unclear
 121         const char chm[] = ":xo,";
 122         char ch = chr[board_ownermap_judge_point(&u->ownermap, c, GJ_THRES)];
 123         if (ch == ',') { // less precise estimate then?
 124                 ch = chm[board_ownermap_judge_point(&u->ownermap, c, 0.67)];
 125         }
 126         fprintf(f, "%c ", ch);
 127 }
 128
 129 static char *
 130 uct_notify_play(struct engine *e, struct board *b, struct move *m)
 131 {
 132         struct uct *u = e->data;
 133         if (!u->t) {
 134                 /* No state, create one - this is probably game beginning
 135                  * and we need to load the opening book right now. */
 136                 prepare_move(e, b, m->color);
 137                 assert(u->t);
 138         }
 139
 140         /* Stop pondering. */
 141         /* XXX: If we are about to receive multiple 'play' commands,
 142          * e.g. in a rengo, we will not ponder during the rest of them. */
 143         uct_pondering_finish(u);
 144
 145         if (is_resign(m->coord)) {
 146                 /* Reset state. */
 147                 reset_state(u);
 148                 return NULL;
 149         }
 150
 151         /* Promote node of the appropriate move to the tree root. */
 152         assert(u->t->root);
 153         if (!tree_promote_at(u->t, b, m->coord)) {
 154                 if (UDEBUGL(0))
 155                         fprintf(stderr, "Warning: Cannot promote move node! Several play commands in row?\n");
 156                 reset_state(u);
 157                 return NULL;
 158         }
 159
 160         return NULL;
 161 }
 162
 163 static char *
 164 uct_chat(struct engine *e, struct board *b, char *cmd)
 165 {
 166         struct uct *u = e->data;
 167         static char reply[1024];
 168
 169         cmd += strspn(cmd, " \n\t");
 170         if (!strncasecmp(cmd, "winrate", 7)) {
 171                 if (!u->t)
 172                         return "no game context (yet?)";
 173                 enum stone color = u->t->root_color;
 174                 struct tree_node *n = u->t->root;
 175                 snprintf(reply, 1024, "In %d*%d playouts, %s %s can win with %.2f%% probability",
 176                          n->u.playouts, u->threads, stone2str(color), coord2sstr(n->coord, b),
 177                          tree_node_get_value(u->t, -1, n->u.value) * 100);
 178                 if (abs(u->t->extra_komi) >= 0.5) {
 179                         sprintf(reply + strlen(reply), ", while self-imposing extra komi %.1f",
 180                                 u->t->extra_komi);
 181                 }
 182                 strcat(reply, ".");
 183                 return reply;
 184         }
 185         return NULL;
 186 }
 187
 188 static void
 189 uct_dead_group_list(struct engine *e, struct board *b, struct move_queue *mq)
 190 {
 191         struct uct *u = e->data;
 192
 193         /* This means the game is probabl yover, no use pondering on. */
 194         uct_pondering_finish(u);
 195
 196         if (u->pass_all_alive)
 197                 return; // no dead groups
 198
 199         bool mock_state = false;
 200
 201         if (!u->t) {
 202                 /* No state, but we cannot just back out - we might
 203                  * have passed earlier, only assuming some stones are
 204                  * dead, and then re-connected, only to lose counting
 205                  * when all stones are assumed alive. */
 206                 /* Mock up some state and seed the ownermap by few
 207                  * simulations. */
 208                 prepare_move(e, b, S_BLACK); assert(u->t);
 209                 for (int i = 0; i < GJ_MINGAMES; i++)
 210                         uct_playout(u, b, S_BLACK, u->t);
 211                 mock_state = true;
 212         }
 213
 214         dead_group_list(u, b, mq);
 215
 216         if (mock_state) {
 217                 /* Clean up the mock state in case we will receive
 218                  * a genmove; we could get a non-alternating-move
 219                  * error from prepare_move() in that case otherwise. */
 220                 reset_state(u);
 221         }
 222 }
 223
 224 static void
 225 playout_policy_done(struct playout_policy *p)
 226 {
 227         if (p->done) p->done(p);
 228         if (p->data) free(p->data);
 229         free(p);
 230 }
 231
 232 static void
 233 uct_done(struct engine *e)
 234 {
 235         /* This is called on engine reset, especially when clear_board
 236          * is received and new game should begin. */
 237         struct uct *u = e->data;
 238         uct_pondering_finish(u);
 239         if (u->t) reset_state(u);
 240         free(u->ownermap.map);
 241
 242         free(u->policy);
 243         free(u->random_policy);
 244         playout_policy_done(u->playout);
 245         uct_prior_done(u->prior);
 246 }
 247
 248
 249 /* Pachi threading structure (if uct_playouts_parallel() is used):
 250  *
 251  * main thread
 252  *   |         main(), GTP communication, ...
 253  *   |
 254  * thread_manager
 255  *   |         spawns and manages worker threads
 256  *   |
 257  * worker0
 258  * worker1
 259  * ...
 260  * workerK
 261  *             uct_playouts() loop, doing descend-playout N=games times
 262  */
 263
 264 /* Set in thread manager in case the workers should stop. */
 265 volatile sig_atomic_t uct_halt = 0;
 266 /* ID of the running worker thread. */
 267 __thread int thread_id = -1;
 268 /* ID of the thread manager. */
 269 static pthread_t thread_manager;
 270 static bool thread_manager_running;
 271
 272 static pthread_mutex_t finish_mutex = PTHREAD_MUTEX_INITIALIZER;
 273 static pthread_cond_t finish_cond = PTHREAD_COND_INITIALIZER;
 274 static volatile int finish_thread;
 275 static pthread_mutex_t finish_serializer = PTHREAD_MUTEX_INITIALIZER;
 276
 277 struct spawn_ctx {
 278         int tid;
 279         struct uct *u;
 280         struct board *b;
 281         enum stone color;
 282         struct tree *t;
 283         unsigned long seed;
 284         int games;
 285 };
 286
 287 static void *
 288 spawn_worker(void *ctx_)
 289 {
 290         struct spawn_ctx *ctx = ctx_;
 291         /* Setup */
 292         fast_srandom(ctx->seed);
 293         thread_id = ctx->tid;
 294         /* Run */
 295         ctx->games = uct_playouts(ctx->u, ctx->b, ctx->color, ctx->t, ctx->games);
 296         /* Finish */
 297         pthread_mutex_lock(&finish_serializer);
 298         pthread_mutex_lock(&finish_mutex);
 299         finish_thread = ctx->tid;
 300         pthread_cond_signal(&finish_cond);
 301         pthread_mutex_unlock(&finish_mutex);
 302         return ctx;
 303 }
 304
 305 /* Thread manager, controlling worker threads. It must be called with
 306  * finish_mutex lock held, and the finish_cond can be signalled for it
 307  * to stop; in that case, the caller should set finish_thread = -1. */
 308 static void *
 309 spawn_thread_manager(void *ctx_)
 310 {
 311         /* In thread_manager, we use only some of the ctx fields. */
 312         struct spawn_ctx *mctx = ctx_;
 313         struct uct *u = mctx->u;
 314         bool shared_tree = u->parallel_tree;
 315         fast_srandom(mctx->seed);
 316
 317         int played_games = 0;
 318         pthread_t threads[u->threads];
 319         int joined = 0;
 320
 321         uct_halt = 0;
 322         /* Spawn threads... */
 323         for (int ti = 0; ti < u->threads; ti++) {
 324                 struct spawn_ctx *ctx = malloc(sizeof(*ctx));
 325                 ctx->u = u; ctx->b = mctx->b; ctx->color = mctx->color;
 326                 ctx->t = shared_tree ? mctx->t : tree_copy(mctx->t);
 327                 ctx->tid = ti; ctx->games = mctx->games;
 328                 ctx->seed = fast_random(65536) + ti;
 329                 pthread_create(&threads[ti], NULL, spawn_worker, ctx);
 330                 if (UDEBUGL(2))
 331                         fprintf(stderr, "Spawned worker %d\n", ti);
 332         }
 333
 334         /* ...and collect them back: */
 335         while (joined < u->threads) {
 336                 /* Wait for some thread to finish... */
 337                 pthread_cond_wait(&finish_cond, &finish_mutex);
 338                 if (finish_thread < 0) {
 339                         /* Stop-by-caller. Tell the workers to wrap up. */
 340                         uct_halt = 1;
 341                         continue;
 342                 }
 343                 /* ...and gather its remnants. */
 344                 struct spawn_ctx *ctx;
 345                 pthread_join(threads[finish_thread], (void **) &ctx);
 346                 played_games += ctx->games;
 347                 joined++;
 348                 if (!shared_tree) {
 349                         tree_merge(mctx->t, ctx->t);
 350                         tree_done(ctx->t);
 351                 }
 352                 free(ctx);
 353                 if (UDEBUGL(2))
 354                         fprintf(stderr, "Joined worker %d\n", finish_thread);
 355                 /* Do not get stalled by slow threads. */
 356                 if (joined >= u->threads / 2)
 357                         uct_halt = 1;
 358                 pthread_mutex_unlock(&finish_serializer);
 359         }
 360
 361         if (!shared_tree)
 362                 tree_normalize(mctx->t, u->threads);
 363
 364         mctx->games = played_games;
 365         return mctx;
 366 }
 367
 368 static void
 369 uct_pondering_start(struct uct *u, struct board *b0, enum stone color, struct tree *t, int games)
 370 {
 371         assert(u->threads > 0);
 372         assert(!thread_manager_running);
 373
 374         /* *b0 can change in the meantime. */
 375         struct board b; board_copy(&b, b0);
 376
 377         struct spawn_ctx ctx = { .u = u, .b = &b, .color = color, .t = t, .games = games, .seed = fast_random(65536) };
 378         static struct spawn_ctx mctx; mctx = ctx;
 379         pthread_mutex_lock(&finish_mutex);
 380         pthread_create(&thread_manager, NULL, spawn_thread_manager, &mctx);
 381         thread_manager_running = true;
 382 }
 383
 384 static int
 385 uct_pondering_stop(void)
 386 {
 387         assert(thread_manager_running);
 388
 389         struct spawn_ctx *pctx;
 390         thread_manager_running = false;
 391         pthread_join(thread_manager, (void **) &pctx);
 392         pthread_mutex_unlock(&finish_mutex);
 393         return pctx->games;
 394 }
 395
 396 /* uct_pondering_stop() frontend for the pondering (non-genmove) mode. */
 397 static void
 398 uct_pondering_finish(struct uct *u)
 399 {
 400         if (!thread_manager_running)
 401                 return;
 402
 403         /* Signal thread manager to stop the workers. */
 404         pthread_mutex_lock(&finish_mutex);
 405         finish_thread = -1;
 406         pthread_cond_signal(&finish_cond);
 407         pthread_mutex_unlock(&finish_mutex);
 408
 409         /* Collect thread manager. */
 410         int games = uct_pondering_stop();
 411         if (UDEBUGL(1))
 412                 fprintf(stderr, "Pondering yielded %d games\n", games);
 413 }
 414
 415 static int
 416 uct_playouts_threaded(struct uct *u, struct board *b, enum stone color, struct tree *t, int games)
 417 {
 418         uct_pondering_start(u, b, color, t, games);
 419         /* We just wait until the thread manager finishes. */
 420         return uct_pondering_stop();
 421 }
 422
 423
 424 static coord_t *
 425 uct_genmove(struct engine *e, struct board *b, enum stone color, bool pass_all_alive)
 426 {
 427         struct uct *u = e->data;
 428
 429         if (b->superko_violation) {
 430                 fprintf(stderr, "!!! WARNING: SUPERKO VIOLATION OCCURED BEFORE THIS MOVE\n");
 431                 fprintf(stderr, "Maybe you play with situational instead of positional superko?\n");
 432                 fprintf(stderr, "I'm going to ignore the violation, but note that I may miss\n");
 433                 fprintf(stderr, "some moves valid under this ruleset because of this.\n");
 434                 b->superko_violation = false;
 435         }
 436
 437         /* Seed the tree. */
 438         uct_pondering_finish(u);
 439         prepare_move(e, b, color);
 440         assert(u->t);
 441
 442         /* Determine number of simulations. */
 443         int games = u->games;
 444         if (u->t->root->children) {
 445                 int delta = u->t->root->u.playouts * 2 / 3;
 446                 if (u->parallel_tree) delta /= u->threads;
 447                 games -= delta;
 448         }
 449         /* else this is highly read-out but dead-end branch of opening book;
 450          * we need to start from scratch; XXX: Maybe actually base the readout
 451          * count based on number of playouts of best node? */
 452         if (games < u->games && UDEBUGL(2))
 453                 fprintf(stderr, "<pre-simulated %d games skipped>\n", u->games - games);
 454
 455         /* Perform the Monte Carlo Tree Search! */
 456         int played_games = uct_playouts_threaded(u, b, color, u->t, games);
 457
 458         if (UDEBUGL(2))
 459                 tree_dump(u->t, u->dumpthres);
 460
 461         /* Choose the best move from the tree. */
 462         struct tree_node *best = u->policy->choose(u->policy, u->t->root, b, color);
 463         if (!best) {
 464                 reset_state(u);
 465                 return coord_copy(pass);
 466         }
 467         if (UDEBUGL(0)) {
 468                 uct_progress_status(u, u->t, color, played_games);
 469         }
 470         if (UDEBUGL(1))
 471                 fprintf(stderr, "*** WINNER is %s (%d,%d) with score %1.4f (%d/%d:%d games)\n",
 472                         coord2sstr(best->coord, b), coord_x(best->coord, b), coord_y(best->coord, b),
 473                         tree_node_get_value(u->t, 1, best->u.value),
 474                         best->u.playouts, u->t->root->u.playouts, played_games);
 475         if (tree_node_get_value(u->t, 1, best->u.value) < u->resign_ratio && !is_pass(best->coord)) {
 476                 reset_state(u);
 477                 return coord_copy(resign);
 478         }
 479
 480         /* If the opponent just passed and we win counting, always
 481          * pass as well. */
 482         if (b->moves > 1 && is_pass(b->last_move.coord)) {
 483                 /* Make sure enough playouts are simulated. */
 484                 while (u->ownermap.playouts < GJ_MINGAMES)
 485                         uct_playout(u, b, color, u->t);
 486                 if (uct_pass_is_safe(u, b, color, u->pass_all_alive || pass_all_alive)) {
 487                         if (UDEBUGL(0))
 488                                 fprintf(stderr, "<Will rather pass, looks safe enough.>\n");
 489                         best->coord = pass;
 490                 }
 491         }
 492
 493         tree_promote_node(u->t, best);
 494         if (u->pondering) {
 495                 if (UDEBUGL(1))
 496                         fprintf(stderr, "Starting to ponder with color %s\n", stone2str(stone_other(color)));
 497                 uct_pondering_start(u, b, stone_other(color), u->t, 0);
 498         }
 499         return coord_copy(best->coord);
 500 }
 501
 502
 503 bool
 504 uct_genbook(struct engine *e, struct board *b, enum stone color)
 505 {
 506         struct uct *u = e->data;
 507         if (!u->t) prepare_move(e, b, color);
 508         assert(u->t);
 509
 510         uct_playouts_threaded(u, b, color, u->t, u->games);
 511
 512         tree_save(u->t, b, u->games / 100);
 513
 514         return true;
 515 }
 516
 517 void
 518 uct_dumpbook(struct engine *e, struct board *b, enum stone color)
 519 {
 520         struct tree *t = tree_init(b, color);
 521         tree_load(t, b);
 522         tree_dump(t, 0);
 523         tree_done(t);
 524 }
 525
 526
 527 struct uct *
 528 uct_state_init(char *arg, struct board *b)
 529 {
 530         struct uct *u = calloc(1, sizeof(struct uct));
 531
 532         u->debug_level = 1;
 533         u->games = MC_GAMES;
 534         u->gamelen = MC_GAMELEN;
 535         u->mercymin = 0;
 536         u->expand_p = 2;
 537         u->dumpthres = 1000;
 538         u->playout_amaf = true;
 539         u->playout_amaf_nakade = false;
 540         u->amaf_prior = false;
 541         u->max_tree_size = 3072ULL * 1048576;
 542
 543         if (board_size(b) - 2 >= 19)
 544                 u->dynkomi = 200;
 545         u->dynkomi_mask = S_BLACK;
 546
 547         u->threads = 1;
 548         u->thread_model = TM_TREEVL;
 549         u->parallel_tree = true;
 550         u->virtual_loss = true;
 551
 552         u->val_scale = 0.02; u->val_points = 20;
 553
 554         if (arg) {
 555                 char *optspec, *next = arg;
 556                 while (*next) {
 557                         optspec = next;
 558                         next += strcspn(next, ",");
 559                         if (*next) { *next++ = 0; } else { *next = 0; }
 560
 561                         char *optname = optspec;
 562                         char *optval = strchr(optspec, '=');
 563                         if (optval) *optval++ = 0;
 564
 565                         if (!strcasecmp(optname, "debug")) {
 566                                 if (optval)
 567                                         u->debug_level = atoi(optval);
 568                                 else
 569                                         u->debug_level++;
 570                         } else if (!strcasecmp(optname, "games") && optval) {
 571                                 u->games = atoi(optval);
 572                         } else if (!strcasecmp(optname, "mercy") && optval) {
 573                                 /* Minimal difference of black/white captures
 574                                  * to stop playout - "Mercy Rule". Speeds up
 575                                  * hopeless playouts at the expense of some
 576                                  * accuracy. */
 577                                 u->mercymin = atoi(optval);
 578                         } else if (!strcasecmp(optname, "gamelen") && optval) {
 579                                 u->gamelen = atoi(optval);
 580                         } else if (!strcasecmp(optname, "expand_p") && optval) {
 581                                 u->expand_p = atoi(optval);
 582                         } else if (!strcasecmp(optname, "dumpthres") && optval) {
 583                                 u->dumpthres = atoi(optval);
 584                         } else if (!strcasecmp(optname, "playout_amaf")) {
 585                                 /* Whether to include random playout moves in
 586                                  * AMAF as well. (Otherwise, only tree moves
 587                                  * are included in AMAF. Of course makes sense
 588                                  * only in connection with an AMAF policy.) */
 589                                 /* with-without: 55.5% (+-4.1) */
 590                                 if (optval && *optval == '0')
 591                                         u->playout_amaf = false;
 592                                 else
 593                                         u->playout_amaf = true;
 594                         } else if (!strcasecmp(optname, "playout_amaf_nakade")) {
 595                                 /* Whether to include nakade moves from playouts
 596                                  * in the AMAF statistics; this tends to nullify
 597                                  * the playout_amaf effect by adding too much
 598                                  * noise. */
 599                                 if (optval && *optval == '0')
 600                                         u->playout_amaf_nakade = false;
 601                                 else
 602                                         u->playout_amaf_nakade = true;
 603                         } else if (!strcasecmp(optname, "playout_amaf_cutoff") && optval) {
 604                                 /* Keep only first N% of playout stage AMAF
 605                                  * information. */
 606                                 u->playout_amaf_cutoff = atoi(optval);
 607                         } else if ((!strcasecmp(optname, "policy") || !strcasecmp(optname, "random_policy")) && optval) {
 608                                 char *policyarg = strchr(optval, ':');
 609                                 struct uct_policy **p = !strcasecmp(optname, "policy") ? &u->policy : &u->random_policy;
 610                                 if (policyarg)
 611                                         *policyarg++ = 0;
 612                                 if (!strcasecmp(optval, "ucb1")) {
 613                                         *p = policy_ucb1_init(u, policyarg);
 614                                 } else if (!strcasecmp(optval, "ucb1amaf")) {
 615                                         *p = policy_ucb1amaf_init(u, policyarg);
 616                                 } else {
 617                                         fprintf(stderr, "UCT: Invalid tree policy %s\n", optval);
 618                                         exit(1);
 619                                 }
 620                         } else if (!strcasecmp(optname, "playout") && optval) {
 621                                 char *playoutarg = strchr(optval, ':');
 622                                 if (playoutarg)
 623                                         *playoutarg++ = 0;
 624                                 if (!strcasecmp(optval, "moggy")) {
 625                                         u->playout = playout_moggy_init(playoutarg);
 626                                 } else if (!strcasecmp(optval, "light")) {
 627                                         u->playout = playout_light_init(playoutarg);
 628                                 } else if (!strcasecmp(optval, "elo")) {
 629                                         u->playout = playout_elo_init(playoutarg);
 630                                 } else {
 631                                         fprintf(stderr, "UCT: Invalid playout policy %s\n", optval);
 632                                         exit(1);
 633                                 }
 634                         } else if (!strcasecmp(optname, "prior") && optval) {
 635                                 u->prior = uct_prior_init(optval, b);
 636                         } else if (!strcasecmp(optname, "amaf_prior") && optval) {
 637                                 u->amaf_prior = atoi(optval);
 638                         } else if (!strcasecmp(optname, "threads") && optval) {
 639                                 /* By default, Pachi will run with only single
 640                                  * tree search thread! */
 641                                 u->threads = atoi(optval);
 642                         } else if (!strcasecmp(optname, "thread_model") && optval) {
 643                                 if (!strcasecmp(optval, "root")) {
 644                                         /* Root parallelization - each thread
 645                                          * does independent search, trees are
 646                                          * merged at the end. */
 647                                         u->thread_model = TM_ROOT;
 648                                         u->parallel_tree = false;
 649                                         u->virtual_loss = false;
 650                                 } else if (!strcasecmp(optval, "tree")) {
 651                                         /* Tree parallelization - all threads
 652                                          * grind on the same tree. */
 653                                         u->thread_model = TM_TREE;
 654                                         u->parallel_tree = true;
 655                                         u->virtual_loss = false;
 656                                 } else if (!strcasecmp(optval, "treevl")) {
 657                                         /* Tree parallelization, but also
 658                                          * with virtual losses - this discou-
 659                                          * rages most threads choosing the
 660                                          * same tree branches to read. */
 661                                         u->thread_model = TM_TREEVL;
 662                                         u->parallel_tree = true;
 663                                         u->virtual_loss = true;
 664                                 } else {
 665                                         fprintf(stderr, "UCT: Invalid thread model %s\n", optval);
 666                                         exit(1);
 667                                 }
 668                         } else if (!strcasecmp(optname, "pondering")) {
 669                                 /* Keep searching even during opponent's turn. */
 670                                 u->pondering = !optval || atoi(optval);
 671                         } else if (!strcasecmp(optname, "force_seed") && optval) {
 672                                 u->force_seed = atoi(optval);
 673                         } else if (!strcasecmp(optname, "no_book")) {
 674                                 u->no_book = true;
 675                         } else if (!strcasecmp(optname, "dynkomi")) {
 676                                 /* Dynamic komi in handicap game; linearly
 677                                  * decreases to basic settings until move
 678                                  * #optval. */
 679                                 u->dynkomi = optval ? atoi(optval) : 150;
 680                         } else if (!strcasecmp(optname, "dynkomi_mask") && optval) {
 681                                 /* Bitmask of colors the player must be
 682                                  * for dynkomi be applied; you may want
 683                                  * to use dynkomi_mask=3 to allow dynkomi
 684                                  * even in games where Pachi is white. */
 685                                 u->dynkomi_mask = atoi(optval);
 686                         } else if (!strcasecmp(optname, "val_scale") && optval) {
 687                                 /* How much of the game result value should be
 688                                  * influenced by win size. Zero means it isn't. */
 689                                 u->val_scale = atof(optval);
 690                         } else if (!strcasecmp(optname, "val_points") && optval) {
 691                                 /* Maximum size of win to be scaled into game
 692                                  * result value. Zero means boardsize^2. */
 693                                 u->val_points = atoi(optval) * 2; // result values are doubled
 694                         } else if (!strcasecmp(optname, "val_extra")) {
 695                                 /* If false, the score coefficient will be simply
 696                                  * added to the value, instead of scaling the result
 697                                  * coefficient because of it. */
 698                                 u->val_extra = !optval || atoi(optval);
 699                         } else if (!strcasecmp(optname, "root_heuristic") && optval) {
 700                                 /* Whether to bias exploration by root node values
 701                                  * (must be supported by the used policy).
 702                                  * 0: Don't.
 703                                  * 1: Do, value = result.
 704                                  * Try to temper the result:
 705                                  * 2: Do, value = 0.5+(result-expected)/2.
 706                                  * 3: Do, value = 0.5+bzz((result-expected)^2). */
 707                                 u->root_heuristic = atoi(optval);
 708                         } else if (!strcasecmp(optname, "pass_all_alive")) {
 709                                 /* Whether to consider all stones alive at the game
 710                                  * end instead of marking dead groupd. */
 711                                 u->pass_all_alive = !optval || atoi(optval);
 712                         } else if (!strcasecmp(optname, "random_policy_chance") && optval) {
 713                                 /* If specified (N), with probability 1/N, random_policy policy
 714                                  * descend is used instead of main policy descend; useful
 715                                  * if specified policy (e.g. UCB1AMAF) can make unduly biased
 716                                  * choices sometimes, you can fall back to e.g.
 717                                  * random_policy=UCB1. */
 718                                 u->random_policy_chance = atoi(optval);
 719                         } else if (!strcasecmp(optname, "max_tree_size") && optval) {
 720                                 /* Maximum amount of memory [MiB] consumed by the move tree.
 721                                  * Default is 3072 (3 GiB). Note that if you use TM_ROOT,
 722                                  * this limits size of only one of the trees, not all of them
 723                                  * together. */
 724                                 u->max_tree_size = atol(optval) * 1048576;
 725                         } else if (!strcasecmp(optname, "banner") && optval) {
 726                                 /* Additional banner string. This must come as the
 727                                  * last engine parameter. */
 728                                 if (*next) *--next = ',';
 729                                 u->banner = strdup(optval);
 730                                 break;
 731                         } else {
 732                                 fprintf(stderr, "uct: Invalid engine argument %s or missing value\n", optname);
 733                                 exit(1);
 734                         }
 735                 }
 736         }
 737
 738         u->resign_ratio = 0.2; /* Resign when most games are lost. */
 739         u->loss_threshold = 0.85; /* Stop reading if after at least 5000 playouts this is best value. */
 740         if (!u->policy)
 741                 u->policy = policy_ucb1amaf_init(u, NULL);
 742
 743         if (!!u->random_policy_chance ^ !!u->random_policy) {
 744                 fprintf(stderr, "uct: Only one of random_policy and random_policy_chance is set\n");
 745                 exit(1);
 746         }
 747
 748         if (!u->prior)
 749                 u->prior = uct_prior_init(NULL, b);
 750
 751         if (!u->playout)
 752                 u->playout = playout_moggy_init(NULL);
 753         u->playout->debug_level = u->debug_level;
 754
 755         u->ownermap.map = malloc(board_size2(b) * sizeof(u->ownermap.map[0]));
 756
 757         /* Some things remain uninitialized for now - the opening book
 758          * is not loaded and the tree not set up. */
 759         /* This will be initialized in setup_state() at the first move
 760          * received/requested. This is because right now we are not aware
 761          * about any komi or handicap setup and such. */
 762
 763         return u;
 764 }
 765
 766 struct engine *
 767 engine_uct_init(char *arg, struct board *b)
 768 {
 769         struct uct *u = uct_state_init(arg, b);
 770         struct engine *e = calloc(1, sizeof(struct engine));
 771         e->name = "UCT Engine";
 772         e->printhook = uct_printhook_ownermap;
 773         e->notify_play = uct_notify_play;
 774         e->chat = uct_chat;
 775         e->genmove = uct_genmove;
 776         e->dead_group_list = uct_dead_group_list;
 777         e->done = uct_done;
 778         e->data = u;
 779
 780         const char banner[] = "I'm playing UCT. When I'm losing, I will resign, "
 781                 "if I think I win, I play until you pass. "
 782                 "Anyone can send me 'winrate' in private chat to get my assessment of the position.";
 783         if (!u->banner) u->banner = "";
 784         e->comment = malloc(sizeof(banner) + strlen(u->banner) + 1);
 785         sprintf(e->comment, "%s %s", banner, u->banner);
 786
 787         return e;
 788 }