uct/uct.c

   1 #include <assert.h>
   2 #include <math.h>
   3 #include <stdio.h>
   4 #include <stdlib.h>
   5 #include <string.h>
   6 #include <time.h>
   7
   8 #define DEBUG
   9
  10 #include "debug.h"
  11 #include "board.h"
  12 #include "gtp.h"
  13 #include "chat.h"
  14 #include "move.h"
  15 #include "mq.h"
  16 #include "dcnn.h"
  17 #include "joseki/base.h"
  18 #include "playout.h"
  19 #include "playout/moggy.h"
  20 #include "playout/light.h"
  21 #include "tactics/util.h"
  22 #include "timeinfo.h"
  23 #include "uct/dynkomi.h"
  24 #include "uct/internal.h"
  25 #include "uct/plugins.h"
  26 #include "uct/prior.h"
  27 #include "uct/search.h"
  28 #include "uct/slave.h"
  29 #include "uct/tree.h"
  30 #include "uct/uct.h"
  31 #include "uct/walk.h"
  32
  33 struct uct_policy *policy_ucb1_init(struct uct *u, char *arg);
  34 struct uct_policy *policy_ucb1amaf_init(struct uct *u, char *arg, struct board *board);
  35 static void uct_pondering_start(struct uct *u, struct board *b0, struct tree *t, enum stone color);
  36
  37 /* Maximal simulation length. */
  38 #define MC_GAMELEN      MAX_GAMELEN
  39
  40
  41 static void
  42 setup_state(struct uct *u, struct board *b, enum stone color)
  43 {
  44         u->t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
  45                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, u->stats_hbits);
  46         if (u->initial_extra_komi)
  47                 u->t->extra_komi = u->initial_extra_komi;
  48         if (u->force_seed)
  49                 fast_srandom(u->force_seed);
  50         if (UDEBUGL(3))
  51                 fprintf(stderr, "Fresh board with random seed %lu\n", fast_getseed());
  52         if (!u->no_tbook && b->moves == 0) {
  53                 if (color == S_BLACK) {
  54                         tree_load(u->t, b);
  55                 } else if (DEBUGL(0)) {
  56                         fprintf(stderr, "Warning: First move appears to be white\n");
  57                 }
  58         }
  59 }
  60
  61 static void
  62 reset_state(struct uct *u)
  63 {
  64         assert(u->t);
  65         tree_done(u->t); u->t = NULL;
  66 }
  67
  68 static void
  69 setup_dynkomi(struct uct *u, struct board *b, enum stone to_play)
  70 {
  71         if (u->t->use_extra_komi && !u->pondering && u->dynkomi->permove)
  72                 u->t->extra_komi = u->dynkomi->permove(u->dynkomi, b, u->t);
  73         else if (!u->t->use_extra_komi)
  74                 u->t->extra_komi = 0;
  75 }
  76
  77 void
  78 uct_prepare_move(struct uct *u, struct board *b, enum stone color)
  79 {
  80         if (u->t) {
  81                 /* Verify that we have sane state. */
  82                 assert(b->es == u);
  83                 assert(u->t && b->moves);
  84                 if (color != stone_other(u->t->root_color)) {
  85                         fprintf(stderr, "Fatal: Non-alternating play detected %d %d\n",
  86                                 color, u->t->root_color);
  87                         exit(1);
  88                 }
  89                 uct_htable_reset(u->t);
  90
  91         } else {
  92                 /* We need fresh state. */
  93                 b->es = u;
  94                 setup_state(u, b, color);
  95         }
  96
  97         u->ownermap.playouts = 0;
  98         memset(u->ownermap.map, 0, board_size2(b) * sizeof(u->ownermap.map[0]));
  99         u->played_own = u->played_all = 0;
 100 }
 101
 102 static void
 103 dead_group_list(struct uct *u, struct board *b, struct move_queue *mq)
 104 {
 105         enum gj_state gs_array[board_size2(b)];
 106         struct group_judgement gj = { .thres = GJ_THRES, .gs = gs_array };
 107         board_ownermap_judge_groups(b, &u->ownermap, &gj);
 108         groups_of_status(b, &gj, GS_DEAD, mq);
 109 }
 110
 111 bool
 112 uct_pass_is_safe(struct uct *u, struct board *b, enum stone color, bool pass_all_alive)
 113 {
 114         /* Make sure enough playouts are simulated to get a reasonable dead group list. */
 115         while (u->ownermap.playouts < GJ_MINGAMES)
 116                 uct_playout(u, b, color, u->t);
 117
 118         struct move_queue mq = { .moves = 0 };
 119         dead_group_list(u, b, &mq);
 120         if (pass_all_alive) {
 121                 for (unsigned int i = 0; i < mq.moves; i++) {
 122                         if (board_at(b, mq.move[i]) == stone_other(color)) {
 123                                 return false; // We need to remove opponent dead groups first.
 124                         }
 125                 }
 126                 mq.moves = 0; // our dead stones are alive when pass_all_alive is true
 127         }
 128         if (u->allow_losing_pass) {
 129                 foreach_point(b) {
 130                         if (board_at(b, c) == S_OFFBOARD)
 131                                 continue;
 132                         if (board_ownermap_judge_point(&u->ownermap, c, GJ_THRES) == PJ_UNKNOWN) {
 133                                 if (UDEBUGL(3))
 134                                         fprintf(stderr, "uct_pass_is_safe fails at %s[%d]\n", coord2sstr(c, b), c);
 135                                 return false; // Unclear point, clarify first.
 136                         }
 137                 } foreach_point_end;
 138                 return true;
 139         }
 140         return pass_is_safe(b, color, &mq);
 141 }
 142
 143 static char *
 144 uct_printhook_ownermap(struct board *board, coord_t c, char *s, char *end)
 145 {
 146         struct uct *u = board->es;
 147         if (!u) {
 148                 strcat(s, ". ");
 149                 return s + 2;
 150         }
 151         const char chr[] = ":XO,"; // dame, black, white, unclear
 152         const char chm[] = ":xo,";
 153         char ch = chr[board_ownermap_judge_point(&u->ownermap, c, GJ_THRES)];
 154         if (ch == ',') { // less precise estimate then?
 155                 ch = chm[board_ownermap_judge_point(&u->ownermap, c, 0.67)];
 156         }
 157         s += snprintf(s, end - s, "%c ", ch);
 158         return s;
 159 }
 160
 161 static float
 162 uct_owner_map(struct engine *e, struct board *b, coord_t c)
 163 {
 164         struct uct *u = b->es;
 165         return board_ownermap_estimate_point(&u->ownermap, c);
 166 }
 167
 168 static char *
 169 uct_notify_play(struct engine *e, struct board *b, struct move *m, char *enginearg)
 170 {
 171         struct uct *u = e->data;
 172         if (!u->t) {
 173                 /* No state, create one - this is probably game beginning
 174                  * and we need to load the opening tbook right now. */
 175                 uct_prepare_move(u, b, m->color);
 176                 assert(u->t);
 177         }
 178
 179         /* Stop pondering, required by tree_promote_at() */
 180         uct_pondering_stop(u);
 181         if (UDEBUGL(2) && u->slave)
 182                 tree_dump(u->t, u->dumpthres);
 183
 184         if (is_resign(m->coord)) {
 185                 /* Reset state. */
 186                 reset_state(u);
 187                 return NULL;
 188         }
 189
 190         /* Promote node of the appropriate move to the tree root. */
 191         assert(u->t->root);
 192         if (u->t->untrustworthy_tree | !tree_promote_at(u->t, b, m->coord)) {
 193                 if (UDEBUGL(3)) {
 194                         if (u->t->untrustworthy_tree)
 195                                 fprintf(stderr, "Not promoting move node in untrustworthy tree.\n");
 196                         else
 197                                 fprintf(stderr, "Warning: Cannot promote move node! Several play commands in row?\n");
 198                 }
 199                 /* Preserve dynamic komi information, though, that is important. */
 200                 u->initial_extra_komi = u->t->extra_komi;
 201                 reset_state(u);
 202                 return NULL;
 203         }
 204
 205         /* If we are a slave in a distributed engine, start pondering once
 206          * we know which move we actually played. See uct_genmove() about
 207          * the check for pass. */
 208         if (u->pondering_opt && u->slave && m->color == u->my_color && !is_pass(m->coord))
 209                 uct_pondering_start(u, b, u->t, stone_other(m->color));
 210
 211         return NULL;
 212 }
 213
 214 static char *
 215 uct_undo(struct engine *e, struct board *b)
 216 {
 217         struct uct *u = e->data;
 218
 219         if (!u->t) return NULL;
 220         uct_pondering_stop(u);
 221         u->initial_extra_komi = u->t->extra_komi;
 222         reset_state(u);
 223         return NULL;
 224 }
 225
 226 static char *
 227 uct_result(struct engine *e, struct board *b)
 228 {
 229         struct uct *u = e->data;
 230         static char reply[1024];
 231
 232         if (!u->t)
 233                 return NULL;
 234         enum stone color = u->t->root_color;
 235         struct tree_node *n = u->t->root;
 236         snprintf(reply, 1024, "%s %s %d %.2f %.1f",
 237                  stone2str(color), coord2sstr(node_coord(n), b),
 238                  n->u.playouts, tree_node_get_value(u->t, -1, n->u.value),
 239                  u->t->use_extra_komi ? u->t->extra_komi : 0);
 240         return reply;
 241 }
 242
 243 static char *
 244 uct_chat(struct engine *e, struct board *b, bool opponent, char *from, char *cmd)
 245 {
 246         struct uct *u = e->data;
 247
 248         if (!u->t)
 249                 return generic_chat(b, opponent, from, cmd, S_NONE, pass, 0, 1, u->threads, 0.0, 0.0);
 250
 251         struct tree_node *n = u->t->root;
 252         double winrate = tree_node_get_value(u->t, -1, n->u.value);
 253         double extra_komi = u->t->use_extra_komi && fabs(u->t->extra_komi) >= 0.5 ? u->t->extra_komi : 0;
 254
 255         return generic_chat(b, opponent, from, cmd, u->t->root_color, node_coord(n), n->u.playouts, 1,
 256                             u->threads, winrate, extra_komi);
 257 }
 258
 259 static void
 260 uct_dead_group_list(struct engine *e, struct board *b, struct move_queue *mq)
 261 {
 262         struct uct *u = e->data;
 263
 264         /* This means the game is probably over, no use pondering on. */
 265         uct_pondering_stop(u);
 266
 267         if (u->pass_all_alive)
 268                 return; // no dead groups
 269
 270         bool mock_state = false;
 271
 272         if (!u->t) {
 273                 /* No state, but we cannot just back out - we might
 274                  * have passed earlier, only assuming some stones are
 275                  * dead, and then re-connected, only to lose counting
 276                  * when all stones are assumed alive. */
 277                 uct_prepare_move(u, b, S_BLACK); assert(u->t);
 278                 mock_state = true;
 279         }
 280         /* Make sure the ownermap is well-seeded. */
 281         while (u->ownermap.playouts < GJ_MINGAMES)
 282                 uct_playout(u, b, S_BLACK, u->t);
 283         /* Show the ownermap: */
 284         if (DEBUGL(2))
 285                 board_print_custom(b, stderr, uct_printhook_ownermap);
 286
 287         dead_group_list(u, b, mq);
 288
 289         if (mock_state) {
 290                 /* Clean up the mock state in case we will receive
 291                  * a genmove; we could get a non-alternating-move
 292                  * error from uct_prepare_move() in that case otherwise. */
 293                 reset_state(u);
 294         }
 295 }
 296
 297 static void
 298 playout_policy_done(struct playout_policy *p)
 299 {
 300         if (p->done) p->done(p);
 301         if (p->data) free(p->data);
 302         free(p);
 303 }
 304
 305 static void
 306 uct_stop(struct engine *e)
 307 {
 308         /* This is called on game over notification. However, an undo
 309          * and game resume can follow, so don't panic yet and just
 310          * relax and stop thinking so that we don't waste CPU. */
 311         struct uct *u = e->data;
 312         uct_pondering_stop(u);
 313 }
 314
 315 static void
 316 uct_done(struct engine *e)
 317 {
 318         /* This is called on engine reset, especially when clear_board
 319          * is received and new game should begin. */
 320         free(e->comment);
 321
 322         struct uct *u = e->data;
 323         uct_pondering_stop(u);
 324         if (u->t) reset_state(u);
 325         if (u->dynkomi) u->dynkomi->done(u->dynkomi);
 326         free(u->ownermap.map);
 327
 328         if (u->policy) u->policy->done(u->policy);
 329         if (u->random_policy) u->random_policy->done(u->random_policy);
 330         playout_policy_done(u->playout);
 331         uct_prior_done(u->prior);
 332         joseki_done(u->jdict);
 333         pluginset_done(u->plugins);
 334 }
 335
 336
 337
 338 /* Run time-limited MCTS search on foreground. */
 339 static int
 340 uct_search(struct uct *u, struct board *b, struct time_info *ti, enum stone color, struct tree *t, bool print_progress)
 341 {
 342         struct uct_search_state s;
 343         uct_search_start(u, b, color, t, ti, &s);
 344         if (UDEBUGL(2) && s.base_playouts > 0)
 345                 fprintf(stderr, "<pre-simulated %d games>\n", s.base_playouts);
 346
 347         /* The search tree is ctx->t. This is currently == . It is important
 348          * to reference ctx->t directly since the
 349          * thread manager will swap the tree pointer asynchronously. */
 350
 351         /* Now, just periodically poll the search tree. */
 352         /* Note that in case of TD_GAMES, threads will not wait for
 353          * the uct_search_check_stop() signalization. */
 354         while (1) {
 355                 time_sleep(TREE_BUSYWAIT_INTERVAL);
 356                 /* TREE_BUSYWAIT_INTERVAL should never be less than desired time, or the
 357                  * time control is broken. But if it happens to be less, we still search
 358                  * at least 100ms otherwise the move is completely random. */
 359
 360                 int i = uct_search_games(&s);
 361                 /* Print notifications etc. */
 362                 uct_search_progress(u, b, color, t, ti, &s, i);
 363                 /* Check if we should stop the search. */
 364                 if (uct_search_check_stop(u, b, color, t, ti, &s, i))
 365                         break;
 366         }
 367
 368         struct uct_thread_ctx *ctx = uct_search_stop();
 369         if (UDEBUGL(2)) tree_dump(t, u->dumpthres);
 370         if (UDEBUGL(2))
 371                 fprintf(stderr, "(avg score %f/%d; dynkomi's %f/%d value %f/%d)\n",
 372                         t->avg_score.value, t->avg_score.playouts,
 373                         u->dynkomi->score.value, u->dynkomi->score.playouts,
 374                         u->dynkomi->value.value, u->dynkomi->value.playouts);
 375         if (print_progress)
 376                 uct_progress_status(u, t, color, ctx->games, NULL);
 377
 378         if (u->debug_after.playouts > 0) {
 379                 /* Now, start an additional run of playouts, single threaded. */
 380                 struct time_info debug_ti = {
 381                         .period = TT_MOVE,
 382                         .dim = TD_GAMES,
 383                 };
 384                 debug_ti.len.games = t->root->u.playouts + u->debug_after.playouts;
 385
 386                 board_print_custom(b, stderr, uct_printhook_ownermap);
 387                 fprintf(stderr, "--8<-- UCT debug post-run begin (%d:%d) --8<--\n", u->debug_after.level, u->debug_after.playouts);
 388
 389                 int debug_level_save = debug_level;
 390                 int u_debug_level_save = u->debug_level;
 391                 int p_debug_level_save = u->playout->debug_level;
 392                 debug_level = u->debug_after.level;
 393                 u->debug_level = u->debug_after.level;
 394                 u->playout->debug_level = u->debug_after.level;
 395                 uct_halt = false;
 396
 397                 uct_playouts(u, b, color, t, &debug_ti);
 398                 tree_dump(t, u->dumpthres);
 399
 400                 uct_halt = true;
 401                 debug_level = debug_level_save;
 402                 u->debug_level = u_debug_level_save;
 403                 u->playout->debug_level = p_debug_level_save;
 404
 405                 fprintf(stderr, "--8<-- UCT debug post-run finished --8<--\n");
 406         }
 407
 408         u->played_own += ctx->games;
 409         return ctx->games;
 410 }
 411
 412 /* Start pondering background with @color to play. */
 413 static void
 414 uct_pondering_start(struct uct *u, struct board *b0, struct tree *t, enum stone color)
 415 {
 416         if (UDEBUGL(1))
 417                 fprintf(stderr, "Starting to ponder with color %s\n", stone2str(stone_other(color)));
 418         u->pondering = true;
 419
 420         /* We need a local board copy to ponder upon. */
 421         struct board *b = malloc2(sizeof(*b)); board_copy(b, b0);
 422
 423         /* *b0 did not have the genmove'd move played yet. */
 424         struct move m = { node_coord(t->root), t->root_color };
 425         int res = board_play(b, &m);
 426         assert(res >= 0);
 427         setup_dynkomi(u, b, stone_other(m.color));
 428
 429         /* Start MCTS manager thread "headless". */
 430         static struct uct_search_state s;
 431         uct_search_start(u, b, color, t, NULL, &s);
 432 }
 433
 434 /* uct_search_stop() frontend for the pondering (non-genmove) mode, and
 435  * to stop the background search for a slave in the distributed engine. */
 436 void
 437 uct_pondering_stop(struct uct *u)
 438 {
 439         if (!thread_manager_running)
 440                 return;
 441
 442         /* Stop the thread manager. */
 443         struct uct_thread_ctx *ctx = uct_search_stop();
 444         if (UDEBUGL(1)) {
 445                 if (u->pondering) fprintf(stderr, "(pondering) ");
 446                 uct_progress_status(u, ctx->t, ctx->color, ctx->games, NULL);
 447         }
 448         if (u->pondering) {
 449                 free(ctx->b);
 450                 u->pondering = false;
 451         }
 452 }
 453
 454
 455 void
 456 uct_genmove_setup(struct uct *u, struct board *b, enum stone color)
 457 {
 458         if (b->superko_violation) {
 459                 fprintf(stderr, "!!! WARNING: SUPERKO VIOLATION OCCURED BEFORE THIS MOVE\n");
 460                 fprintf(stderr, "Maybe you play with situational instead of positional superko?\n");
 461                 fprintf(stderr, "I'm going to ignore the violation, but note that I may miss\n");
 462                 fprintf(stderr, "some moves valid under this ruleset because of this.\n");
 463                 b->superko_violation = false;
 464         }
 465
 466         uct_prepare_move(u, b, color);
 467
 468         assert(u->t);
 469         u->my_color = color;
 470
 471         /* How to decide whether to use dynkomi in this game? Since we use
 472          * pondering, it's not simple "who-to-play" matter. Decide based on
 473          * the last genmove issued. */
 474         u->t->use_extra_komi = !!(u->dynkomi_mask & color);
 475         setup_dynkomi(u, b, color);
 476
 477         if (b->rules == RULES_JAPANESE)
 478                 u->territory_scoring = true;
 479
 480         /* Make pessimistic assumption about komi for Japanese rules to
 481          * avoid losing by 0.5 when winning by 0.5 with Chinese rules.
 482          * The rules usually give the same winner if the integer part of komi
 483          * is odd so we adjust the komi only if it is even (for a board of
 484          * odd size). We are not trying  to get an exact evaluation for rare
 485          * cases of seki. For details see http://home.snafu.de/jasiek/parity.html */
 486         if (u->territory_scoring && (((int)floor(b->komi) + board_size(b)) & 1)) {
 487                 b->komi += (color == S_BLACK ? 1.0 : -1.0);
 488                 if (UDEBUGL(0))
 489                         fprintf(stderr, "Setting komi to %.1f assuming Japanese rules\n",
 490                                 b->komi);
 491         }
 492 }
 493
 494 static void
 495 uct_live_gfx_hook(struct engine *e)
 496 {
 497         struct uct *u = e->data;
 498         /* Hack: Override reportfreq to get decent update rates in GoGui */
 499         u->reportfreq = 1000;
 500 }
 501
 502 /* Kindof like uct_genmove() but just find the best candidates */
 503 static void
 504 uct_best_moves(struct engine *e, struct board *b, enum stone color)
 505 {
 506         struct time_info ti = { .period = TT_NULL };
 507         double start_time = time_now();
 508         struct uct *u = e->data;
 509         uct_pondering_stop(u);
 510         if (u->t)
 511                 reset_state(u);
 512         uct_genmove_setup(u, b, color);
 513
 514         /* Start the Monte Carlo Tree Search! */
 515         int base_playouts = u->t->root->u.playouts;
 516         int played_games = uct_search(u, b, &ti, color, u->t, false);
 517
 518         coord_t best_coord;
 519         uct_search_result(u, b, color, u->pass_all_alive, played_games, base_playouts, &best_coord);
 520
 521         if (UDEBUGL(2)) {
 522                 double time = time_now() - start_time + 0.000001; /* avoid divide by zero */
 523                 fprintf(stderr, "genmove in %0.2fs (%d games/s, %d games/s/thread)\n",
 524                         time, (int)(played_games/time), (int)(played_games/time/u->threads));
 525         }
 526
 527         uct_progress_status(u, u->t, color, played_games, &best_coord);
 528         reset_state(u);
 529 }
 530
 531 static coord_t *
 532 uct_genmove(struct engine *e, struct board *b, struct time_info *ti, enum stone color, bool pass_all_alive)
 533 {
 534         double start_time = time_now();
 535         struct uct *u = e->data;
 536         u->pass_all_alive |= pass_all_alive;
 537         uct_pondering_stop(u);
 538         uct_genmove_setup(u, b, color);
 539
 540         /* Start the Monte Carlo Tree Search! */
 541         int base_playouts = u->t->root->u.playouts;
 542         int played_games = uct_search(u, b, ti, color, u->t, false);
 543
 544         coord_t best_coord;
 545         struct tree_node *best;
 546         best = uct_search_result(u, b, color, u->pass_all_alive, played_games, base_playouts, &best_coord);
 547
 548         if (UDEBUGL(2)) {
 549                 double time = time_now() - start_time + 0.000001; /* avoid divide by zero */
 550                 fprintf(stderr, "genmove in %0.2fs (%d games/s, %d games/s/thread)\n",
 551                         time, (int)(played_games/time), (int)(played_games/time/u->threads));
 552         }
 553
 554         uct_progress_status(u, u->t, color, played_games, &best_coord);
 555
 556         if (!best) {
 557                 /* Pass or resign. */
 558                 if (is_pass(best_coord))
 559                         u->initial_extra_komi = u->t->extra_komi;
 560                 reset_state(u);
 561                 return coord_copy(best_coord);
 562         }
 563
 564         if (!u->t->untrustworthy_tree) {
 565                 tree_promote_node(u->t, &best);
 566         } else {
 567                 /* Throw away an untrustworthy tree. */
 568                 /* Preserve dynamic komi information, though, that is important. */
 569                 u->initial_extra_komi = u->t->extra_komi;
 570                 reset_state(u);
 571         }
 572
 573         /* After a pass, pondering is harmful for two reasons:
 574          * (i) We might keep pondering even when the game is over.
 575          * Of course this is the case for opponent resign as well.
 576          * (ii) More importantly, the ownermap will get skewed since
 577          * the UCT will start cutting off any playouts. */
 578         if (u->pondering_opt && u->t && !is_pass(node_coord(best))) {
 579                 uct_pondering_start(u, b, u->t, stone_other(color));
 580         }
 581         return coord_copy(best_coord);
 582 }
 583
 584
 585 bool
 586 uct_gentbook(struct engine *e, struct board *b, struct time_info *ti, enum stone color)
 587 {
 588         struct uct *u = e->data;
 589         if (!u->t) uct_prepare_move(u, b, color);
 590         assert(u->t);
 591
 592         if (ti->dim == TD_GAMES) {
 593                 /* Don't count in games that already went into the tbook. */
 594                 ti->len.games += u->t->root->u.playouts;
 595         }
 596         uct_search(u, b, ti, color, u->t, true);
 597
 598         assert(ti->dim == TD_GAMES);
 599         tree_save(u->t, b, ti->len.games / 100);
 600
 601         return true;
 602 }
 603
 604 void
 605 uct_dumptbook(struct engine *e, struct board *b, enum stone color)
 606 {
 607         struct uct *u = e->data;
 608         struct tree *t = tree_init(b, color, u->fast_alloc ? u->max_tree_size : 0,
 609                          u->max_pruned_size, u->pruning_threshold, u->local_tree_aging, 0);
 610         tree_load(t, b);
 611         tree_dump(t, 0);
 612         tree_done(t);
 613 }
 614
 615
 616 floating_t
 617 uct_evaluate_one(struct engine *e, struct board *b, struct time_info *ti, coord_t c, enum stone color)
 618 {
 619         struct uct *u = e->data;
 620
 621         struct board b2;
 622         board_copy(&b2, b);
 623         struct move m = { c, color };
 624         int res = board_play(&b2, &m);
 625         if (res < 0)
 626                 return NAN;
 627         color = stone_other(color);
 628
 629         if (u->t) reset_state(u);
 630         uct_prepare_move(u, &b2, color);
 631         assert(u->t);
 632
 633         floating_t bestval;
 634         uct_search(u, &b2, ti, color, u->t, true);
 635         struct tree_node *best = u->policy->choose(u->policy, u->t->root, &b2, color, resign);
 636         if (!best) {
 637                 bestval = NAN; // the opponent has no reply!
 638         } else {
 639                 bestval = tree_node_get_value(u->t, 1, best->u.value);
 640         }
 641
 642         reset_state(u); // clean our junk
 643
 644         return isnan(bestval) ? NAN : 1.0f - bestval;
 645 }
 646
 647 void
 648 uct_evaluate(struct engine *e, struct board *b, struct time_info *ti, floating_t *vals, enum stone color)
 649 {
 650         for (int i = 0; i < b->flen; i++) {
 651                 if (is_pass(b->f[i]))
 652                         vals[i] = NAN;
 653                 else
 654                         vals[i] = uct_evaluate_one(e, b, ti, b->f[i], color);
 655         }
 656 }
 657
 658
 659 struct uct *
 660 uct_state_init(char *arg, struct board *b)
 661 {
 662         struct uct *u = calloc2(1, sizeof(struct uct));
 663         bool pat_setup = false;
 664
 665         u->debug_level = debug_level;
 666         u->reportfreq = 10000;
 667         u->gamelen = MC_GAMELEN;
 668         u->resign_threshold = 0.2;
 669         u->sure_win_threshold = 0.95;
 670         u->mercymin = 0;
 671         u->significant_threshold = 50;
 672         u->expand_p = 8;
 673         u->dumpthres = 0.01;
 674         u->playout_amaf = true;
 675         u->amaf_prior = false;
 676         u->max_tree_size = 1408ULL * 1048576;
 677         u->fast_alloc = true;
 678         u->pruning_threshold = 0;
 679
 680         u->threads = 1;
 681         u->thread_model = TM_TREEVL;
 682         u->virtual_loss = 1;
 683
 684         u->pondering_opt = true;
 685
 686         u->fuseki_end = 20; // max time at 361*20% = 72 moves (our 36th move, still 99 to play)
 687         u->yose_start = 40; // (100-40-25)*361/100/2 = 63 moves still to play by us then
 688         u->bestr_ratio = 0.02;
 689         // 2.5 is clearly too much, but seems to compensate well for overly stern time allocations.
 690         // TODO: Further tuning and experiments with better time allocation schemes.
 691         u->best2_ratio = 2.5;
 692         // Higher values of max_maintime_ratio sometimes cause severe time trouble in tournaments
 693         // It might be necessary to reduce it to 1.5 on large board, but more tuning is needed.
 694         u->max_maintime_ratio = 2.0;
 695
 696         u->val_scale = 0; u->val_points = 40;
 697         u->dynkomi_interval = 1000;
 698         u->dynkomi_mask = S_BLACK | S_WHITE;
 699
 700         u->tenuki_d = 4;
 701         u->local_tree_aging = 80;
 702         u->local_tree_depth_decay = 1.5;
 703         u->local_tree_eval = LTE_ROOT;
 704         u->local_tree_neival = true;
 705
 706         u->max_slaves = -1;
 707         u->slave_index = -1;
 708         u->stats_delay = 0.01; // 10 ms
 709         u->shared_levels = 1;
 710
 711         u->plugins = pluginset_init(b);
 712
 713         u->jdict = joseki_load(b->size);
 714
 715         if (arg) {
 716                 char *optspec, *next = arg;
 717                 while (*next) {
 718                         optspec = next;
 719                         next += strcspn(next, ",");
 720                         if (*next) { *next++ = 0; } else { *next = 0; }
 721
 722                         char *optname = optspec;
 723                         char *optval = strchr(optspec, '=');
 724                         if (optval) *optval++ = 0;
 725
 726                         /** Basic options */
 727
 728                         if (!strcasecmp(optname, "debug")) {
 729                                 if (optval)
 730                                         u->debug_level = atoi(optval);
 731                                 else
 732                                         u->debug_level++;
 733                         } else if (!strcasecmp(optname, "reporting") && optval) {
 734                                 /* The format of output for detailed progress
 735                                  * information (such as current best move and
 736                                  * its value, etc.). */
 737                                 if (!strcasecmp(optval, "text")) {
 738                                         /* Plaintext traditional output. */
 739                                         u->reporting = UR_TEXT;
 740                                 } else if (!strcasecmp(optval, "json")) {
 741                                         /* JSON output. Implies debug=0. */
 742                                         u->reporting = UR_JSON;
 743                                         u->debug_level = 0;
 744                                 } else if (!strcasecmp(optval, "jsonbig")) {
 745                                         /* JSON output, but much more detailed.
 746                                          * Implies debug=0. */
 747                                         u->reporting = UR_JSON_BIG;
 748                                         u->debug_level = 0;
 749                                 } else {
 750                                         fprintf(stderr, "UCT: Invalid reporting format %s\n", optval);
 751                                         exit(1);
 752                                 }
 753                         } else if (!strcasecmp(optname, "reportfreq") && optval) {
 754                                 /* The progress information line will be shown
 755                                  * every <reportfreq> simulations. */
 756                                 u->reportfreq = atoi(optval);
 757                         } else if (!strcasecmp(optname, "dumpthres") && optval) {
 758                                 /* When dumping the UCT tree on output, include
 759                                  * nodes with at least this many playouts.
 760                                  * (A fraction of the total # of playouts at the
 761                                  * tree root.) */
 762                                 /* Use 0 to list all nodes with at least one
 763                                  * simulation, and -1 to list _all_ nodes. */
 764                                 u->dumpthres = atof(optval);
 765                         } else if (!strcasecmp(optname, "resign_threshold") && optval) {
 766                                 /* Resign when this ratio of games is lost
 767                                  * after GJ_MINGAMES sample is taken. */
 768                                 u->resign_threshold = atof(optval);
 769                         } else if (!strcasecmp(optname, "sure_win_threshold") && optval) {
 770                                 /* Stop reading when this ratio of games is won
 771                                  * after PLAYOUT_EARLY_BREAK_MIN sample is
 772                                  * taken. (Prevents stupid time losses,
 773                                  * friendly to human opponents.) */
 774                                 u->sure_win_threshold = atof(optval);
 775                         } else if (!strcasecmp(optname, "force_seed") && optval) {
 776                                 /* Set RNG seed at the tree setup. */
 777                                 u->force_seed = atoi(optval);
 778                         } else if (!strcasecmp(optname, "no_tbook")) {
 779                                 /* Disable UCT opening tbook. */
 780                                 u->no_tbook = true;
 781                         } else if (!strcasecmp(optname, "pass_all_alive")) {
 782                                 /* Whether to consider passing only after all
 783                                  * dead groups were removed from the board;
 784                                  * this is like all genmoves are in fact
 785                                  * kgs-genmove_cleanup. */
 786                                 u->pass_all_alive = !optval || atoi(optval);
 787                         } else if (!strcasecmp(optname, "allow_losing_pass")) {
 788                                 /* Whether to consider passing in a clear
 789                                  * but losing situation, to be scored as a loss
 790                                  * for us. */
 791                                 u->allow_losing_pass = !optval || atoi(optval);
 792                         } else if (!strcasecmp(optname, "territory_scoring")) {
 793                                 /* Use territory scoring (default is area scoring).
 794                                  * An explicit kgs-rules command overrides this. */
 795                                 u->territory_scoring = !optval || atoi(optval);
 796                         } else if (!strcasecmp(optname, "stones_only")) {
 797                                 /* Do not count eyes. Nice to teach go to kids.
 798                                  * http://strasbourg.jeudego.org/regle_strasbourgeoise.htm */
 799                                 b->rules = RULES_STONES_ONLY;
 800                                 u->pass_all_alive = true;
 801                         } else if (!strcasecmp(optname, "debug_after")) {
 802                                 /* debug_after=9:1000 will make Pachi think under
 803                                  * the normal conditions, but at the point when
 804                                  * a move is to be chosen, the tree is dumped and
 805                                  * another 1000 simulations are run single-threaded
 806                                  * with debug level 9, allowing inspection of Pachi's
 807                                  * behavior after it has thought a lot. */
 808                                 if (optval) {
 809                                         u->debug_after.level = atoi(optval);
 810                                         char *playouts = strchr(optval, ':');
 811                                         if (playouts)
 812                                                 u->debug_after.playouts = atoi(playouts+1);
 813                                         else
 814                                                 u->debug_after.playouts = 1000;
 815                                 } else {
 816                                         u->debug_after.level = 9;
 817                                         u->debug_after.playouts = 1000;
 818                                 }
 819                         } else if (!strcasecmp(optname, "banner") && optval) {
 820                                 /* Additional banner string. This must come as the
 821                                  * last engine parameter. You can use '+' instead
 822                                  * of ' ' if you are wrestling with kgsGtp. */
 823                                 if (*next) *--next = ',';
 824                                 u->banner = strdup(optval);
 825                                 for (char *b = u->banner; *b; b++) {
 826                                         if (*b == '+') *b = ' ';
 827                                 }
 828                                 break;
 829                         } else if (!strcasecmp(optname, "plugin") && optval) {
 830                                 /* Load an external plugin; filename goes before the colon,
 831                                  * extra arguments after the colon. */
 832                                 char *pluginarg = strchr(optval, ':');
 833                                 if (pluginarg)
 834                                         *pluginarg++ = 0;
 835                                 plugin_load(u->plugins, optval, pluginarg);
 836
 837                         /** UCT behavior and policies */
 838
 839                         } else if ((!strcasecmp(optname, "policy")
 840                                 /* Node selection policy. ucb1amaf is the
 841                                  * default policy implementing RAVE, while
 842                                  * ucb1 is the simple exploration/exploitation
 843                                  * policy. Policies can take further extra
 844                                  * options. */
 845                                     || !strcasecmp(optname, "random_policy")) && optval) {
 846                                 /* A policy to be used randomly with small
 847                                  * chance instead of the default policy. */
 848                                 char *policyarg = strchr(optval, ':');
 849                                 struct uct_policy **p = !strcasecmp(optname, "policy") ? &u->policy : &u->random_policy;
 850                                 if (policyarg)
 851                                         *policyarg++ = 0;
 852                                 if (!strcasecmp(optval, "ucb1")) {
 853                                         *p = policy_ucb1_init(u, policyarg);
 854                                 } else if (!strcasecmp(optval, "ucb1amaf")) {
 855                                         *p = policy_ucb1amaf_init(u, policyarg, b);
 856                                 } else {
 857                                         fprintf(stderr, "UCT: Invalid tree policy %s\n", optval);
 858                                         exit(1);
 859                                 }
 860                         } else if (!strcasecmp(optname, "playout") && optval) {
 861                                 /* Random simulation (playout) policy.
 862                                  * moggy is the default policy with large
 863                                  * amount of domain-specific knowledge and
 864                                  * heuristics. light is a simple uniformly
 865                                  * random move selection policy. */
 866                                 char *playoutarg = strchr(optval, ':');
 867                                 if (playoutarg)
 868                                         *playoutarg++ = 0;
 869                                 if (!strcasecmp(optval, "moggy")) {
 870                                         u->playout = playout_moggy_init(playoutarg, b, u->jdict);
 871                                 } else if (!strcasecmp(optval, "light")) {
 872                                         u->playout = playout_light_init(playoutarg, b);
 873                                 } else {
 874                                         fprintf(stderr, "UCT: Invalid playout policy %s\n", optval);
 875                                         exit(1);
 876                                 }
 877                         } else if (!strcasecmp(optname, "prior") && optval) {
 878                                 /* Node priors policy. When expanding a node,
 879                                  * it will seed node values heuristically
 880                                  * (most importantly, based on playout policy
 881                                  * opinion, but also with regard to other
 882                                  * things). See uct/prior.c for details.
 883                                  * Use prior=eqex=0 to disable priors. */
 884                                 u->prior = uct_prior_init(optval, b, u);
 885                         } else if (!strcasecmp(optname, "mercy") && optval) {
 886                                 /* Minimal difference of black/white captures
 887                                  * to stop playout - "Mercy Rule". Speeds up
 888                                  * hopeless playouts at the expense of some
 889                                  * accuracy. */
 890                                 u->mercymin = atoi(optval);
 891                         } else if (!strcasecmp(optname, "gamelen") && optval) {
 892                                 /* Maximum length of single simulation
 893                                  * in moves. */
 894                                 u->gamelen = atoi(optval);
 895                         } else if (!strcasecmp(optname, "expand_p") && optval) {
 896                                 /* Expand UCT nodes after it has been
 897                                  * visited this many times. */
 898                                 u->expand_p = atoi(optval);
 899                         } else if (!strcasecmp(optname, "random_policy_chance") && optval) {
 900                                 /* If specified (N), with probability 1/N, random_policy policy
 901                                  * descend is used instead of main policy descend; useful
 902                                  * if specified policy (e.g. UCB1AMAF) can make unduly biased
 903                                  * choices sometimes, you can fall back to e.g.
 904                                  * random_policy=UCB1. */
 905                                 u->random_policy_chance = atoi(optval);
 906
 907                         /** General AMAF behavior */
 908                         /* (Only relevant if the policy supports AMAF.
 909                          * More variables can be tuned as policy
 910                          * parameters.) */
 911
 912                         } else if (!strcasecmp(optname, "playout_amaf")) {
 913                                 /* Whether to include random playout moves in
 914                                  * AMAF as well. (Otherwise, only tree moves
 915                                  * are included in AMAF. Of course makes sense
 916                                  * only in connection with an AMAF policy.) */
 917                                 /* with-without: 55.5% (+-4.1) */
 918                                 if (optval && *optval == '0')
 919                                         u->playout_amaf = false;
 920                                 else
 921                                         u->playout_amaf = true;
 922                         } else if (!strcasecmp(optname, "playout_amaf_cutoff") && optval) {
 923                                 /* Keep only first N% of playout stage AMAF
 924                                  * information. */
 925                                 u->playout_amaf_cutoff = atoi(optval);
 926                         } else if (!strcasecmp(optname, "amaf_prior") && optval) {
 927                                 /* In node policy, consider prior values
 928                                  * part of the real result term or part
 929                                  * of the AMAF term? */
 930                                 u->amaf_prior = atoi(optval);
 931
 932                         /** Performance and memory management */
 933
 934                         } else if (!strcasecmp(optname, "threads") && optval) {
 935                                 /* By default, Pachi will run with only single
 936                                  * tree search thread! */
 937                                 u->threads = atoi(optval);
 938                         } else if (!strcasecmp(optname, "thread_model") && optval) {
 939                                 if (!strcasecmp(optval, "tree")) {
 940                                         /* Tree parallelization - all threads
 941                                          * grind on the same tree. */
 942                                         u->thread_model = TM_TREE;
 943                                         u->virtual_loss = 0;
 944                                 } else if (!strcasecmp(optval, "treevl")) {
 945                                         /* Tree parallelization, but also
 946                                          * with virtual losses - this discou-
 947                                          * rages most threads choosing the
 948                                          * same tree branches to read. */
 949                                         u->thread_model = TM_TREEVL;
 950                                 } else {
 951                                         fprintf(stderr, "UCT: Invalid thread model %s\n", optval);
 952                                         exit(1);
 953                                 }
 954                         } else if (!strcasecmp(optname, "virtual_loss") && optval) {
 955                                 /* Number of virtual losses added before evaluating a node. */
 956                                 u->virtual_loss = atoi(optval);
 957                         } else if (!strcasecmp(optname, "pondering")) {
 958                                 /* Keep searching even during opponent's turn. */
 959                                 u->pondering_opt = !optval || atoi(optval);
 960                         } else if (!strcasecmp(optname, "max_tree_size") && optval) {
 961                                 /* Maximum amount of memory [MiB] consumed by the move tree.
 962                                  * For fast_alloc it includes the temp tree used for pruning.
 963                                  * Default is 3072 (3 GiB). */
 964                                 u->max_tree_size = atol(optval) * 1048576;
 965                         } else if (!strcasecmp(optname, "fast_alloc")) {
 966                                 u->fast_alloc = !optval || atoi(optval);
 967                         } else if (!strcasecmp(optname, "pruning_threshold") && optval) {
 968                                 /* Force pruning at beginning of a move if the tree consumes
 969                                  * more than this [MiB]. Default is 10% of max_tree_size.
 970                                  * Increase to reduce pruning time overhead if memory is plentiful.
 971                                  * This option is meaningful only for fast_alloc. */
 972                                 u->pruning_threshold = atol(optval) * 1048576;
 973
 974                         /** Time control */
 975
 976                         } else if (!strcasecmp(optname, "best2_ratio") && optval) {
 977                                 /* If set, prolong simulating while
 978                                  * first_best/second_best playouts ratio
 979                                  * is less than best2_ratio. */
 980                                 u->best2_ratio = atof(optval);
 981                         } else if (!strcasecmp(optname, "bestr_ratio") && optval) {
 982                                 /* If set, prolong simulating while
 983                                  * best,best_best_child values delta
 984                                  * is more than bestr_ratio. */
 985                                 u->bestr_ratio = atof(optval);
 986                         } else if (!strcasecmp(optname, "max_maintime_ratio") && optval) {
 987                                 /* If set and while not in byoyomi, prolong simulating no more than
 988                                  * max_maintime_ratio times the normal desired thinking time. */
 989                                 u->max_maintime_ratio = atof(optval);
 990                         } else if (!strcasecmp(optname, "fuseki_end") && optval) {
 991                                 /* At the very beginning it's not worth thinking
 992                                  * too long because the playout evaluations are
 993                                  * very noisy. So gradually increase the thinking
 994                                  * time up to maximum when fuseki_end percent
 995                                  * of the board has been played.
 996                                  * This only applies if we are not in byoyomi. */
 997                                 u->fuseki_end = atoi(optval);
 998                         } else if (!strcasecmp(optname, "yose_start") && optval) {
 999                                 /* When yose_start percent of the board has been
1000                                  * played, or if we are in byoyomi, stop spending
1001                                  * more time and spread the remaining time
1002                                  * uniformly.
1003                                  * Between fuseki_end and yose_start, we spend
1004                                  * a constant proportion of the remaining time
1005                                  * on each move. (yose_start should actually
1006                                  * be much earlier than when real yose start,
1007                                  * but "yose" is a good short name to convey
1008                                  * the idea.) */
1009                                 u->yose_start = atoi(optval);
1010
1011                         /** Dynamic komi */
1012
1013                         } else if (!strcasecmp(optname, "dynkomi") && optval) {
1014                                 /* Dynamic komi approach; there are multiple
1015                                  * ways to adjust komi dynamically throughout
1016                                  * play. We currently support two: */
1017                                 char *dynkomiarg = strchr(optval, ':');
1018                                 if (dynkomiarg)
1019                                         *dynkomiarg++ = 0;
1020                                 if (!strcasecmp(optval, "none")) {
1021                                         u->dynkomi = uct_dynkomi_init_none(u, dynkomiarg, b);
1022                                 } else if (!strcasecmp(optval, "linear")) {
1023                                         /* You should set dynkomi_mask=1 or a very low
1024                                          * handicap_value for white. */
1025                                         u->dynkomi = uct_dynkomi_init_linear(u, dynkomiarg, b);
1026                                 } else if (!strcasecmp(optval, "adaptive")) {
1027                                         /* There are many more knobs to
1028                                          * crank - see uct/dynkomi.c. */
1029                                         u->dynkomi = uct_dynkomi_init_adaptive(u, dynkomiarg, b);
1030                                 } else {
1031                                         fprintf(stderr, "UCT: Invalid dynkomi mode %s\n", optval);
1032                                         exit(1);
1033                                 }
1034                         } else if (!strcasecmp(optname, "dynkomi_mask") && optval) {
1035                                 /* Bitmask of colors the player must be
1036                                  * for dynkomi be applied; the default dynkomi_mask=3 allows
1037                                  * dynkomi even in games where Pachi is white. */
1038                                 u->dynkomi_mask = atoi(optval);
1039                         } else if (!strcasecmp(optname, "dynkomi_interval") && optval) {
1040                                 /* If non-zero, re-adjust dynamic komi
1041                                  * throughout a single genmove reading,
1042                                  * roughly every N simulations. */
1043                                 /* XXX: Does not work with tree
1044                                  * parallelization. */
1045                                 u->dynkomi_interval = atoi(optval);
1046                         } else if (!strcasecmp(optname, "extra_komi") && optval) {
1047                                 /* Initial dynamic komi settings. This
1048                                  * is useful for the adaptive dynkomi
1049                                  * policy as the value to start with
1050                                  * (this is NOT kept fixed) in case
1051                                  * there is not enough time in the search
1052                                  * to adjust the value properly (e.g. the
1053                                  * game was interrupted). */
1054                                 u->initial_extra_komi = atof(optval);
1055
1056                         /** Node value result scaling */
1057
1058                         } else if (!strcasecmp(optname, "val_scale") && optval) {
1059                                 /* How much of the game result value should be
1060                                  * influenced by win size. Zero means it isn't. */
1061                                 u->val_scale = atof(optval);
1062                         } else if (!strcasecmp(optname, "val_points") && optval) {
1063                                 /* Maximum size of win to be scaled into game
1064                                  * result value. Zero means boardsize^2. */
1065                                 u->val_points = atoi(optval) * 2; // result values are doubled
1066                         } else if (!strcasecmp(optname, "val_extra")) {
1067                                 /* If false, the score coefficient will be simply
1068                                  * added to the value, instead of scaling the result
1069                                  * coefficient because of it. */
1070                                 u->val_extra = !optval || atoi(optval);
1071                         } else if (!strcasecmp(optname, "val_byavg")) {
1072                                 /* If true, the score included in the value will
1073                                  * be relative to average score in the current
1074                                  * search episode inst. of jigo. */
1075                                 u->val_byavg = !optval || atoi(optval);
1076                         } else if (!strcasecmp(optname, "val_bytemp")) {
1077                                 /* If true, the value scaling coefficient
1078                                  * is different based on value extremity
1079                                  * (dist. from 0.5), linear between
1080                                  * val_bytemp_min, val_scale. */
1081                                 u->val_bytemp = !optval || atoi(optval);
1082                         } else if (!strcasecmp(optname, "val_bytemp_min") && optval) {
1083                                 /* Minimum val_scale in case of val_bytemp. */
1084                                 u->val_bytemp_min = atof(optval);
1085
1086                         /** Local trees */
1087                         /* (Purely experimental. Does not work - yet!) */
1088
1089                         } else if (!strcasecmp(optname, "local_tree")) {
1090                                 /* Whether to bias exploration by local tree values. */
1091                                 u->local_tree = !optval || atoi(optval);
1092                         } else if (!strcasecmp(optname, "tenuki_d") && optval) {
1093                                 /* Tenuki distance at which to break the local tree. */
1094                                 u->tenuki_d = atoi(optval);
1095                                 if (u->tenuki_d > TREE_NODE_D_MAX + 1) {
1096                                         fprintf(stderr, "uct: tenuki_d must not be larger than TREE_NODE_D_MAX+1 %d\n", TREE_NODE_D_MAX + 1);
1097                                         exit(1);
1098                                 }
1099                         } else if (!strcasecmp(optname, "local_tree_aging") && optval) {
1100                                 /* How much to reduce local tree values between moves. */
1101                                 u->local_tree_aging = atof(optval);
1102                         } else if (!strcasecmp(optname, "local_tree_depth_decay") && optval) {
1103                                 /* With value x>0, during the descent the node
1104                                  * contributes 1/x^depth playouts in
1105                                  * the local tree. I.e., with x>1, nodes more
1106                                  * distant from local situation contribute more
1107                                  * than nodes near the root. */
1108                                 u->local_tree_depth_decay = atof(optval);
1109                         } else if (!strcasecmp(optname, "local_tree_allseq")) {
1110                                 /* If disabled, only complete sequences are stored
1111                                  * in the local tree. If this is on, also
1112                                  * subsequences starting at each move are stored. */
1113                                 u->local_tree_allseq = !optval || atoi(optval);
1114                         } else if (!strcasecmp(optname, "local_tree_neival")) {
1115                                 /* If disabled, local node value is not
1116                                  * computed just based on terminal status
1117                                  * of the coordinate, but also its neighbors. */
1118                                 u->local_tree_neival = !optval || atoi(optval);
1119                         } else if (!strcasecmp(optname, "local_tree_eval")) {
1120                                 /* How is the value inserted in the local tree
1121                                  * determined. */
1122                                 if (!strcasecmp(optval, "root"))
1123                                         /* All moves within a tree branch are
1124                                          * considered wrt. their merit
1125                                          * reaching tachtical goal of making
1126                                          * the first move in the branch
1127                                          * survive. */
1128                                         u->local_tree_eval = LTE_ROOT;
1129                                 else if (!strcasecmp(optval, "each"))
1130                                         /* Each move is considered wrt.
1131                                          * its own survival. */
1132                                         u->local_tree_eval = LTE_EACH;
1133                                 else if (!strcasecmp(optval, "total"))
1134                                         /* The tactical goal is the survival
1135                                          * of all the moves of my color and
1136                                          * non-survival of all the opponent
1137                                          * moves. Local values (and their
1138                                          * inverses) are averaged. */
1139                                         u->local_tree_eval = LTE_TOTAL;
1140                                 else {
1141                                         fprintf(stderr, "uct: unknown local_tree_eval %s\n", optval);
1142                                         exit(1);
1143                                 }
1144                         } else if (!strcasecmp(optname, "local_tree_rootchoose")) {
1145                                 /* If disabled, only moves within the local
1146                                  * tree branch are considered; the values
1147                                  * of the branch roots (i.e. root children)
1148                                  * are ignored. This may make sense together
1149                                  * with eval!=each, we consider only moves
1150                                  * that influence the goal, not the "rating"
1151                                  * of the goal itself. (The real solution
1152                                  * will be probably using criticality to pick
1153                                  * local tree branches.) */
1154                                 u->local_tree_rootchoose = !optval || atoi(optval);
1155
1156                         /** Other heuristics */
1157                         } else if (!strcasecmp(optname, "patterns")) {
1158                                 /* Load pattern database. Various modules
1159                                  * (priors, policies etc.) may make use
1160                                  * of this database. They will request
1161                                  * it automatically in that case, but you
1162                                  * can use this option to tweak the pattern
1163                                  * parameters. */
1164                                 patterns_init(&u->pat, optval, false, true);
1165                                 u->want_pat = pat_setup = true;
1166                         } else if (!strcasecmp(optname, "significant_threshold") && optval) {
1167                                 /* Some heuristics (XXX: none in mainline) rely
1168                                  * on the knowledge of the last "significant"
1169                                  * node in the descent. Such a node is
1170                                  * considered reasonably trustworthy to carry
1171                                  * some meaningful information in the values
1172                                  * of the node and its children. */
1173                                 u->significant_threshold = atoi(optval);
1174
1175                         /** Distributed engine slaves setup */
1176
1177                         } else if (!strcasecmp(optname, "slave")) {
1178                                 /* Act as slave for the distributed engine. */
1179                                 u->slave = !optval || atoi(optval);
1180                         } else if (!strcasecmp(optname, "slave_index") && optval) {
1181                                 /* Optional index if per-slave behavior is desired.
1182                                  * Must be given as index/max */
1183                                 u->slave_index = atoi(optval);
1184                                 char *p = strchr(optval, '/');
1185                                 if (p) u->max_slaves = atoi(++p);
1186                         } else if (!strcasecmp(optname, "shared_nodes") && optval) {
1187                                 /* Share at most shared_nodes between master and slave at each genmoves.
1188                                  * Must use the same value in master and slaves. */
1189                                 u->shared_nodes = atoi(optval);
1190                         } else if (!strcasecmp(optname, "shared_levels") && optval) {
1191                                 /* Share only nodes of level <= shared_levels. */
1192                                 u->shared_levels = atoi(optval);
1193                         } else if (!strcasecmp(optname, "stats_hbits") && optval) {
1194                                 /* Set hash table size to 2^stats_hbits for the shared stats. */
1195                                 u->stats_hbits = atoi(optval);
1196                         } else if (!strcasecmp(optname, "stats_delay") && optval) {
1197                                 /* How long to wait in slave for initial stats to build up before
1198                                  * replying to the genmoves command (in ms) */
1199                                 u->stats_delay = 0.001 * atof(optval);
1200
1201                         /** Presets */
1202
1203                         } else if (!strcasecmp(optname, "maximize_score")) {
1204                                 /* A combination of settings that will make
1205                                  * Pachi try to maximize his points (instead
1206                                  * of playing slack yose) or minimize his loss
1207                                  * (and proceed to counting even when losing). */
1208                                 /* Please note that this preset might be
1209                                  * somewhat weaker than normal Pachi, and the
1210                                  * score maximization is approximate; point size
1211                                  * of win/loss still should not be used to judge
1212                                  * strength of Pachi or the opponent. */
1213                                 /* See README for some further notes. */
1214                                 if (!optval || atoi(optval)) {
1215                                         /* Allow scoring a lost game. */
1216                                         u->allow_losing_pass = true;
1217                                         /* Make Pachi keep his calm when losing
1218                                          * and/or maintain winning marging. */
1219                                         /* Do not play games that are losing
1220                                          * by too much. */
1221                                         /* XXX: komi_ratchet_age=40000 is necessary
1222                                          * with losing_komi_ratchet, but 40000
1223                                          * is somewhat arbitrary value. */
1224                                         char dynkomi_args[] = "losing_komi_ratchet:komi_ratchet_age=60000:no_komi_at_game_end=0:max_losing_komi=30";
1225                                         u->dynkomi = uct_dynkomi_init_adaptive(u, dynkomi_args, b);
1226                                         /* XXX: Values arbitrary so far. */
1227                                         /* XXX: Also, is bytemp sensible when
1228                                          * combined with dynamic komi?! */
1229                                         u->val_scale = 0.01;
1230                                         u->val_bytemp = true;
1231                                         u->val_bytemp_min = 0.001;
1232                                         u->val_byavg = true;
1233                                 }
1234
1235                         } else {
1236                                 fprintf(stderr, "uct: Invalid engine argument %s or missing value\n", optname);
1237                                 exit(1);
1238                         }
1239                 }
1240         }
1241
1242         if (!u->policy)
1243                 u->policy = policy_ucb1amaf_init(u, NULL, b);
1244
1245         if (!!u->random_policy_chance ^ !!u->random_policy) {
1246                 fprintf(stderr, "uct: Only one of random_policy and random_policy_chance is set\n");
1247                 exit(1);
1248         }
1249
1250         if (!u->local_tree) {
1251                 /* No ltree aging. */
1252                 u->local_tree_aging = 1.0f;
1253         }
1254
1255         if (u->fast_alloc) {
1256                 if (u->pruning_threshold < u->max_tree_size / 10)
1257                         u->pruning_threshold = u->max_tree_size / 10;
1258                 if (u->pruning_threshold > u->max_tree_size / 2)
1259                         u->pruning_threshold = u->max_tree_size / 2;
1260
1261                 /* Limit pruning temp space to 20% of memory. Beyond this we discard
1262                  * the nodes and recompute them at the next move if necessary. */
1263                 u->max_pruned_size = u->max_tree_size / 5;
1264                 u->max_tree_size -= u->max_pruned_size;
1265         } else {
1266                 /* Reserve 5% memory in case the background free() are slower
1267                  * than the concurrent allocations. */
1268                 u->max_tree_size -= u->max_tree_size / 20;
1269         }
1270
1271         if (!u->prior)
1272                 u->prior = uct_prior_init(NULL, b, u);
1273
1274         if (!u->playout)
1275                 u->playout = playout_moggy_init(NULL, b, u->jdict);
1276         if (!u->playout->debug_level)
1277                 u->playout->debug_level = u->debug_level;
1278
1279         if (u->want_pat && !pat_setup)
1280                 patterns_init(&u->pat, NULL, false, true);
1281         dcnn_init();
1282
1283         u->ownermap.map = malloc2(board_size2(b) * sizeof(u->ownermap.map[0]));
1284
1285         if (u->slave) {
1286                 if (!u->stats_hbits) u->stats_hbits = DEFAULT_STATS_HBITS;
1287                 if (!u->shared_nodes) u->shared_nodes = DEFAULT_SHARED_NODES;
1288                 assert(u->shared_levels * board_bits2(b) <= 8 * (int)sizeof(path_t));
1289         }
1290
1291         if (!u->dynkomi)
1292                 u->dynkomi = board_small(b) ? uct_dynkomi_init_none(u, NULL, b)
1293                         : uct_dynkomi_init_linear(u, NULL, b);
1294
1295         /* Some things remain uninitialized for now - the opening tbook
1296          * is not loaded and the tree not set up. */
1297         /* This will be initialized in setup_state() at the first move
1298          * received/requested. This is because right now we are not aware
1299          * about any komi or handicap setup and such. */
1300
1301         return u;
1302 }
1303
1304 struct engine *
1305 engine_uct_init(char *arg, struct board *b)
1306 {
1307         struct uct *u = uct_state_init(arg, b);
1308         struct engine *e = calloc2(1, sizeof(struct engine));
1309         e->name = "UCT";
1310         e->printhook = uct_printhook_ownermap;
1311         e->notify_play = uct_notify_play;
1312         e->chat = uct_chat;
1313         e->undo = uct_undo;
1314         e->result = uct_result;
1315         e->genmove = uct_genmove;
1316         e->genmoves = uct_genmoves;
1317         e->evaluate = uct_evaluate;
1318         e->dead_group_list = uct_dead_group_list;
1319         e->stop = uct_stop;
1320         e->done = uct_done;
1321         e->owner_map = uct_owner_map;
1322         e->best_moves = uct_best_moves;
1323         e->live_gfx_hook = uct_live_gfx_hook;
1324         e->data = u;
1325         if (u->slave)
1326                 e->notify = uct_notify;
1327
1328         const char banner[] = "If you believe you have won but I am still playing, "
1329                 "please help me understand by capturing all dead stones. "
1330                 "Anyone can send me 'winrate' in private chat to get my assessment of the position.";
1331         if (!u->banner) u->banner = "";
1332         e->comment = malloc2(sizeof(banner) + strlen(u->banner) + 1);
1333         sprintf(e->comment, "%s %s", banner, u->banner);
1334
1335         return e;
1336 }