uct/prior.c

   1 #include <assert.h>
   2 #include <math.h>
   3 #include <stdio.h>
   4 #include <stdlib.h>
   5 #include <string.h>
   6
   7 #include "board.h"
   8 #include "debug.h"
   9 #include "joseki/base.h"
  10 #include "move.h"
  11 #include "random.h"
  12 #include "tactics/ladder.h"
  13 #include "tactics/util.h"
  14 #include "uct/internal.h"
  15 #include "uct/plugins.h"
  16 #include "uct/prior.h"
  17 #include "uct/tree.h"
  18
  19 /* Applying heuristic values to the tree nodes, skewing the reading in
  20  * most interesting directions. */
  21
  22 /* TODO: Introduce foreach_fpoint() to iterate only over non-occupied
  23  * positions. */
  24
  25 struct uct_prior {
  26         /* Equivalent experience for prior knowledge. MoGo paper recommends
  27          * 50 playouts per source; in practice, esp. with RAVE, about 6
  28          * playouts per source seems best. */
  29         int eqex;
  30         int even_eqex, policy_eqex, b19_eqex, eye_eqex, ko_eqex, plugin_eqex, joseki_eqex, pattern_eqex;
  31         int cfgdn; int *cfgd_eqex;
  32         bool prune_ladders;
  33 };
  34
  35 void
  36 uct_prior_even(struct uct *u, struct tree_node *node, struct prior_map *map)
  37 {
  38         /* Q_{even} */
  39         /* This may be dubious for normal UCB1 but is essential for
  40          * reading stability of RAVE, it appears. */
  41         add_prior_value(map, pass, 0.5, u->prior->even_eqex);
  42         foreach_free_point(map->b) {
  43                 if (!map->consider[c])
  44                         continue;
  45                 add_prior_value(map, c, 0.5, u->prior->even_eqex);
  46         } foreach_free_point_end;
  47 }
  48
  49 void
  50 uct_prior_eye(struct uct *u, struct tree_node *node, struct prior_map *map)
  51 {
  52         /* Discourage playing into our own eyes. However, we cannot
  53          * completely prohibit it:
  54          * #######
  55          * ...XX.#
  56          * XOOOXX#
  57          * X.OOOO#
  58          * .XXXX.# */
  59         foreach_free_point(map->b) {
  60                 if (!map->consider[c])
  61                         continue;
  62                 if (!board_is_one_point_eye(map->b, c, map->to_play))
  63                         continue;
  64                 add_prior_value(map, c, 0, u->prior->eye_eqex);
  65         } foreach_free_point_end;
  66 }
  67
  68 void
  69 uct_prior_ko(struct uct *u, struct tree_node *node, struct prior_map *map)
  70 {
  71         /* Favor fighting ko, if we took it le 10 moves ago. */
  72         coord_t ko = map->b->last_ko.coord;
  73         if (is_pass(ko) || map->b->moves - map->b->last_ko_age > 10 || !map->consider[ko])
  74                 return;
  75         // fprintf(stderr, "prior ko-fight @ %s %s\n", stone2str(map->to_play), coord2sstr(ko, map->b));
  76         add_prior_value(map, ko, 1, u->prior->ko_eqex);
  77 }
  78
  79 void
  80 uct_prior_b19(struct uct *u, struct tree_node *node, struct prior_map *map)
  81 {
  82         /* Q_{b19} */
  83         /* Specific hints for 19x19 board - priors for certain edge distances. */
  84         foreach_free_point(map->b) {
  85                 if (!map->consider[c])
  86                         continue;
  87                 int d = coord_edge_distance(c, map->b);
  88                 if (d != 0 && d != 2)
  89                         continue;
  90                 /* The bonus applies only with no stones in immediate
  91                  * vincinity. */
  92                 if (board_stone_radar(map->b, c, 2))
  93                         continue;
  94                 /* First line: 0 */
  95                 /* Third line: 1 */
  96                 add_prior_value(map, c, d == 2, u->prior->b19_eqex);
  97         } foreach_free_point_end;
  98 }
  99
 100 void
 101 uct_prior_playout(struct uct *u, struct tree_node *node, struct prior_map *map)
 102 {
 103         /* Q_{playout-policy} */
 104         if (u->playout->assess)
 105                 u->playout->assess(u->playout, map, u->prior->policy_eqex);
 106 }
 107
 108 void
 109 uct_prior_cfgd(struct uct *u, struct tree_node *node, struct prior_map *map)
 110 {
 111         /* Q_{common_fate_graph_distance} */
 112         /* Give bonus to moves local to the last move, where "local" means
 113          * local in terms of groups, not just manhattan distance. */
 114         if (is_pass(map->b->last_move.coord) || is_resign(map->b->last_move.coord))
 115                 return;
 116
 117         foreach_free_point(map->b) {
 118                 if (!map->consider[c])
 119                         continue;
 120                 if (map->distances[c] > u->prior->cfgdn)
 121                         continue;
 122                 assert(map->distances[c] != 0);
 123                 int bonus = u->prior->cfgd_eqex[map->distances[c]];
 124                 add_prior_value(map, c, 1, bonus);
 125         } foreach_free_point_end;
 126 }
 127
 128 void
 129 uct_prior_joseki(struct uct *u, struct tree_node *node, struct prior_map *map)
 130 {
 131         /* Q_{joseki} */
 132         if (!u->jdict)
 133                 return;
 134         for (int i = 0; i < 4; i++) {
 135                 hash_t h = map->b->qhash[i] & joseki_hash_mask;
 136                 coord_t *cc = u->jdict->patterns[h].moves[map->to_play - 1];
 137                 if (!cc) continue;
 138                 for (; !is_pass(*cc); cc++) {
 139                         if (coord_quadrant(*cc, map->b) != i)
 140                                 continue;
 141                         add_prior_value(map, *cc, 1.0, u->prior->joseki_eqex);
 142                 }
 143         }
 144 }
 145
 146 void
 147 uct_prior_pattern(struct uct *u, struct tree_node *node, struct prior_map *map)
 148 {
 149         /* Q_{pattern} */
 150         if (!u->pat.pd)
 151                 return;
 152
 153         struct board *b = map->b;
 154         struct pattern pats[b->flen];
 155         floating_t probs[b->flen];
 156         pattern_rate_moves(&u->pat, b, map->to_play, pats, probs);
 157         if (UDEBUGL(5)) {
 158                 fprintf(stderr, "Pattern prior at node %s\n", coord2sstr(node->coord, b));
 159                 board_print(b, stderr);
 160         }
 161
 162         for (int f = 0; f < b->flen; f++) {
 163                 if (isnan(probs[f]) || probs[f] < 0.001)
 164                         continue;
 165                 assert(!is_pass(b->f[f]));
 166                 if (UDEBUGL(5)) {
 167                         char s[256]; pattern2str(s, &pats[f]);
 168                         fprintf(stderr, "\t%s: %.3f %s\n", coord2sstr(b->f[f], b), probs[f], s);
 169                 }
 170                 add_prior_value(map, b->f[f], 1.0, sqrt(probs[f]) * u->prior->pattern_eqex);
 171         }
 172 }
 173
 174 void
 175 uct_prior(struct uct *u, struct tree_node *node, struct prior_map *map)
 176 {
 177         if (u->prior->prune_ladders) {
 178                 foreach_free_point(map->b) {
 179                         if (!map->consider[c])
 180                                 continue;
 181                         group_t atari_neighbor = board_get_atari_neighbor(map->b, c, map->to_play);
 182                         if (atari_neighbor && is_ladder(map->b, c, atari_neighbor, true))
 183                                 map->consider[c] = false;
 184                 } foreach_free_point_end;
 185         }
 186
 187         if (u->prior->even_eqex)
 188                 uct_prior_even(u, node, map);
 189         if (u->prior->eye_eqex)
 190                 uct_prior_eye(u, node, map);
 191         if (u->prior->ko_eqex)
 192                 uct_prior_ko(u, node, map);
 193         if (u->prior->b19_eqex)
 194                 uct_prior_b19(u, node, map);
 195         if (u->prior->policy_eqex)
 196                 uct_prior_playout(u, node, map);
 197         if (u->prior->cfgd_eqex)
 198                 uct_prior_cfgd(u, node, map);
 199         if (u->prior->joseki_eqex)
 200                 uct_prior_joseki(u, node, map);
 201         if (u->prior->pattern_eqex)
 202                 uct_prior_pattern(u, node, map);
 203         if (u->prior->plugin_eqex)
 204                 plugin_prior(u->plugins, node, map, u->prior->plugin_eqex);
 205 }
 206
 207 struct uct_prior *
 208 uct_prior_init(char *arg, struct board *b, struct uct *u)
 209 {
 210         struct uct_prior *p = calloc2(1, sizeof(struct uct_prior));
 211
 212         p->even_eqex = p->policy_eqex = p->b19_eqex = p->eye_eqex = p->ko_eqex = p->plugin_eqex = -100;
 213         /* FIXME: Optimal pattern_eqex is about -1000 with small playout counts
 214          * but only -400 on a cluster. We need a better way to set the default
 215          * here. */
 216         p->pattern_eqex = -400;
 217         p->joseki_eqex = -200;
 218         p->cfgdn = -1;
 219
 220         /* Even number! */
 221         p->eqex = board_large(b) ? 20 : 14;
 222
 223         if (arg) {
 224                 char *optspec, *next = arg;
 225                 while (*next) {
 226                         optspec = next;
 227                         next += strcspn(next, ":");
 228                         if (*next) { *next++ = 0; } else { *next = 0; }
 229
 230                         char *optname = optspec;
 231                         char *optval = strchr(optspec, '=');
 232                         if (optval) *optval++ = 0;
 233
 234                         if (!strcasecmp(optname, "eqex") && optval) {
 235                                 p->eqex = atoi(optval);
 236
 237                         /* In the following settings, you can use negative
 238                          * numbers to give the hundredths of default eqex.
 239                          * E.g. -100 is default eqex, -50 is half of the
 240                          * default eqex, -200 is double the default eqex. */
 241                         } else if (!strcasecmp(optname, "even") && optval) {
 242                                 p->even_eqex = atoi(optval);
 243                         } else if (!strcasecmp(optname, "policy") && optval) {
 244                                 p->policy_eqex = atoi(optval);
 245                         } else if (!strcasecmp(optname, "b19") && optval) {
 246                                 p->b19_eqex = atoi(optval);
 247                         } else if (!strcasecmp(optname, "cfgd") && optval) {
 248                                 /* cfgd=3%40%20%20 - 3 levels; immediate libs
 249                                  * of last move => 40 wins, their neighbors
 250                                  * 20 wins, 2nd-level neighbors 20 wins;
 251                                  * neighbors are group-transitive. */
 252                                 p->cfgdn = atoi(optval); optval += strcspn(optval, "%");
 253                                 p->cfgd_eqex = calloc2(p->cfgdn + 1, sizeof(*p->cfgd_eqex));
 254                                 p->cfgd_eqex[0] = 0;
 255                                 int i;
 256                                 for (i = 1; *optval; i++, optval += strcspn(optval, "%")) {
 257                                         optval++;
 258                                         p->cfgd_eqex[i] = atoi(optval);
 259                                 }
 260                                 if (i != p->cfgdn + 1) {
 261                                         fprintf(stderr, "uct: Missing prior cfdn level %d/%d\n", i, p->cfgdn);
 262                                         exit(1);
 263                                 }
 264
 265                         } else if (!strcasecmp(optname, "joseki") && optval) {
 266                                 p->joseki_eqex = atoi(optval);
 267                         } else if (!strcasecmp(optname, "eye") && optval) {
 268                                 p->eye_eqex = atoi(optval);
 269                         } else if (!strcasecmp(optname, "ko") && optval) {
 270                                 p->ko_eqex = atoi(optval);
 271                         } else if (!strcasecmp(optname, "pattern") && optval) {
 272                                 /* Pattern-based prior eqex. */
 273                                 /* Note that this prior is still going to be
 274                                  * used only if you have downloaded or
 275                                  * generated the pattern files! */
 276                                 p->pattern_eqex = atoi(optval);
 277                         } else if (!strcasecmp(optname, "plugin") && optval) {
 278                                 /* Unlike others, this is just a *recommendation*. */
 279                                 p->plugin_eqex = atoi(optval);
 280                         } else if (!strcasecmp(optname, "prune_ladders")) {
 281                                 p->prune_ladders = !optval || atoi(optval);
 282                         } else {
 283                                 fprintf(stderr, "uct: Invalid prior argument %s or missing value\n", optname);
 284                                 exit(1);
 285                         }
 286                 }
 287         }
 288
 289         if (p->even_eqex < 0) p->even_eqex = p->eqex * -p->even_eqex / 100;
 290         if (p->policy_eqex < 0) p->policy_eqex = p->eqex * -p->policy_eqex / 100;
 291         if (p->b19_eqex < 0) p->b19_eqex = p->eqex * -p->b19_eqex / 100;
 292         if (p->eye_eqex < 0) p->eye_eqex = p->eqex * -p->eye_eqex / 100;
 293         if (p->ko_eqex < 0) p->ko_eqex = p->eqex * -p->ko_eqex / 100;
 294         if (p->joseki_eqex < 0) p->joseki_eqex = p->eqex * -p->joseki_eqex / 100;
 295         if (p->pattern_eqex < 0) p->pattern_eqex = p->eqex * -p->pattern_eqex / 100;
 296         if (p->plugin_eqex < 0) p->plugin_eqex = p->eqex * -p->plugin_eqex / 100;
 297
 298         if (p->cfgdn < 0) {
 299                 static int large_bonuses[] = { 0, 55, 50, 15 };
 300                 static int small_bonuses[] = { 0, 45, 40, 15 };
 301                 p->cfgdn = 3;
 302                 p->cfgd_eqex = calloc2(p->cfgdn + 1, sizeof(*p->cfgd_eqex));
 303                 memcpy(p->cfgd_eqex, board_large(b) ? large_bonuses : small_bonuses, sizeof(large_bonuses));
 304         }
 305         if (p->cfgdn > TREE_NODE_D_MAX) {
 306                 fprintf(stderr, "uct: CFG distances only up to %d available\n", TREE_NODE_D_MAX);
 307                 exit(1);
 308         }
 309
 310         if (p->pattern_eqex)
 311                 u->want_pat = true;
 312
 313         return p;
 314 }
 315
 316 void
 317 uct_prior_done(struct uct_prior *p)
 318 {
 319         assert(p->cfgd_eqex);
 320         free(p->cfgd_eqex);
 321         free(p);
 322 }