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