playout.c

   1 #define DEBUG
   2 #include <assert.h>
   3 #include <math.h>
   4 #include <stdio.h>
   5 #include <stdlib.h>
   6 #include <string.h>
   7
   8 #include "board.h"
   9 #include "debug.h"
  10 #include "engine.h"
  11 #include "move.h"
  12 #include "ownermap.h"
  13 #include "playout.h"
  14 #include "tactics/goals.h"
  15
  16 /* Whether to set global debug level to the same as the playout
  17  * has, in case it is different. This can make sure e.g. tactical
  18  * reading produces proper level of debug prints during simulations.
  19  * But it is safe to enable this only in single-threaded instances! */
  20 //#define DEBUGL_BY_PLAYOUT
  21
  22 #define PLDEBUGL(n) DEBUGL_(policy->debug_level, n)
  23
  24
  25 coord_t
  26 play_random_move(struct playout_setup *setup,
  27                  struct board *b, enum stone color,
  28                  struct playout_policy *policy)
  29 {
  30         coord_t coord = pass;
  31
  32         if (setup->prepolicy_hook) {
  33                 coord = setup->prepolicy_hook(policy, setup, b, color);
  34                 // fprintf(stderr, "prehook: %s\n", coord2sstr(coord, b));
  35         }
  36
  37         if (is_pass(coord)) {
  38                 coord = policy->choose(policy, setup, b, color);
  39                 // fprintf(stderr, "policy: %s\n", coord2sstr(coord, b));
  40         }
  41
  42         if (is_pass(coord) && setup->postpolicy_hook) {
  43                 coord = setup->postpolicy_hook(policy, setup, b, color);
  44                 // fprintf(stderr, "posthook: %s\n", coord2sstr(coord, b));
  45         }
  46
  47         if (is_pass(coord)) {
  48 play_random:
  49                 /* Defer to uniformly random move choice. */
  50                 /* This must never happen if the policy is tracking
  51                  * internal board state, obviously. */
  52                 assert(!policy->setboard);
  53                 board_play_random(b, color, &coord, (ppr_permit) policy->permit, policy);
  54
  55         } else {
  56                 struct move m;
  57                 m.coord = coord; m.color = color;
  58                 if (board_play(b, &m) < 0) {
  59                         if (PLDEBUGL(4)) {
  60                                 fprintf(stderr, "Pre-picked move %d,%d is ILLEGAL:\n",
  61                                         coord_x(coord, b), coord_y(coord, b));
  62                                 board_print(b, stderr);
  63                         }
  64                         goto play_random;
  65                 }
  66         }
  67
  68         return coord;
  69 }
  70
  71 int
  72 play_random_game(struct playout_setup *setup,
  73                  struct board *b, enum stone starting_color,
  74                  struct playout_amafmap *amafmap,
  75                  struct board_ownermap *ownermap,
  76                  struct playout_policy *policy)
  77 {
  78         assert(setup && policy);
  79
  80         int gamelen = setup->gamelen - b->moves;
  81
  82         struct libmap_mq lmqueue = {{0}};
  83         if (b->libmap) {
  84                 b->lmqueue = &lmqueue;
  85                 b->libmap_init_groups = false;
  86         }
  87
  88         if (policy->setboard)
  89                 policy->setboard(policy, b);
  90 #ifdef DEBUGL_BY_PLAYOUT
  91         int debug_level_orig = debug_level;
  92         debug_level = policy->debug_level;
  93 #endif
  94
  95         enum stone color = starting_color;
  96
  97         int passes = is_pass(b->last_move.coord) && b->moves > 0;
  98
  99         while (gamelen-- && passes < 2) {
 100                 coord_t coord = play_random_move(setup, b, color, policy);
 101
 102 #if 0
 103                 /* For UCT, superko test here is downright harmful since
 104                  * in superko-likely situation we throw away literally
 105                  * 95% of our playouts; UCT will deal with this fine by
 106                  * itself. */
 107                 if (unlikely(b->superko_violation)) {
 108                         /* We ignore superko violations that are suicides. These
 109                          * are common only at the end of the game and are
 110                          * rather harmless. (They will not go through as a root
 111                          * move anyway.) */
 112                         if (group_at(b, coord)) {
 113                                 if (DEBUGL(3)) {
 114                                         fprintf(stderr, "Superko fun at %d,%d in\n", coord_x(coord, b), coord_y(coord, b));
 115                                         if (DEBUGL(4))
 116                                                 board_print(b, stderr);
 117                                 }
 118                                 return 0;
 119                         } else {
 120                                 if (DEBUGL(6)) {
 121                                         fprintf(stderr, "Ignoring superko at %d,%d in\n", coord_x(coord, b), coord_y(coord, b));
 122                                         board_print(b, stderr);
 123                                 }
 124                                 b->superko_violation = false;
 125                         }
 126                 }
 127 #endif
 128
 129                 if (PLDEBUGL(7)) {
 130                         fprintf(stderr, "%s %s\n", stone2str(color), coord2sstr(coord, b));
 131                         if (PLDEBUGL(8))
 132                                 board_print(b, stderr);
 133                 }
 134
 135                 if (unlikely(is_pass(coord))) {
 136                         passes++;
 137                 } else {
 138                         passes = 0;
 139                 }
 140                 if (amafmap) {
 141                         assert(amafmap->gamelen < MAX_GAMELEN);
 142                         amafmap->is_ko_capture[amafmap->gamelen] = board_playing_ko_threat(b);
 143                         amafmap->game[amafmap->gamelen++] = coord;
 144                 }
 145
 146                 if (setup->mercymin && abs(b->captures[S_BLACK] - b->captures[S_WHITE]) > setup->mercymin)
 147                         break;
 148
 149                 color = stone_other(color);
 150         }
 151
 152         floating_t score = board_fast_score(b);
 153         int result = (starting_color == S_WHITE ? score * 2 : - (score * 2));
 154
 155         if (DEBUGL(6)) {
 156                 fprintf(stderr, "Random playout result: %d (W %f)\n", result, score);
 157                 if (DEBUGL(7))
 158                         board_print(b, stderr);
 159         }
 160
 161         if (ownermap)
 162                 board_ownermap_fill(ownermap, b);
 163         if (b->libmap) {
 164                 libmap_queue_process(b, score > 0 ? S_WHITE : S_BLACK);
 165                 b->lmqueue = NULL;
 166         }
 167
 168         if (b->ps)
 169                 free(b->ps);
 170
 171 #ifdef DEBUGL_BY_PLAYOUT
 172         debug_level = debug_level_orig;
 173 #endif
 174
 175         return result;
 176 }