uct/policy/generic.c

   1 #include <assert.h>
   2 #include <math.h>
   3 #include <stdio.h>
   4 #include <stdlib.h>
   5
   6 #include "board.h"
   7 #include "debug.h"
   8 #include "move.h"
   9 #include "tactics/util.h"
  10 #include "random.h"
  11 #include "uct/internal.h"
  12 #include "uct/tree.h"
  13 #include "uct/policy/generic.h"
  14
  15 struct tree_node *
  16 uctp_generic_choose(struct uct_policy *p, struct tree_node *node, struct board *b, enum stone color, coord_t exclude)
  17 {
  18         struct tree_node *nbest = NULL;
  19         /* This function is called while the tree is updated by other threads.
  20          * We rely on node->children being set only after the node has been fully expanded. */
  21         for (struct tree_node *ni = node->children; ni; ni = ni->sibling)
  22                 // we compare playouts and choose the best-explored
  23                 // child; comparing values is more brittle
  24                 if (!nbest || ni->u.playouts > nbest->u.playouts) {
  25                         if (node_coord(ni) == exclude)
  26                                 continue;
  27                         if (ni->hints & TREE_HINT_INVALID)
  28                                 continue;
  29                         /* Play pass only if we can afford scoring */
  30                         if (is_pass(node_coord(ni)) && !uct_pass_is_safe(p->uct, b, color, p->uct->pass_all_alive))
  31                                 continue;
  32                         nbest = ni;
  33                 }
  34         return nbest;
  35 }
  36
  37 /* Return the node with best value instead of best explored. We must use the heuristic
  38  * value (using prior and possibly rave), because the raw value is meaningless for
  39  * nodes evaluated rarely.
  40  * This function is called while the tree is updated by other threads */
  41 void
  42 uctp_generic_winner(struct uct_policy *p, struct tree *tree, struct uct_descent *descent)
  43 {
  44         if (!p->evaluate)
  45                 return;
  46         bool allow_pass = false; /* At worst forces some extra playouts at the end */
  47         int parity = tree_node_parity(tree, descent->node);
  48
  49         uctd_try_node_children(tree, descent, allow_pass, parity, p->uct->tenuki_d, di, urgency) {
  50                 urgency = p->evaluate(p, tree, &di, parity);
  51         } uctd_set_best_child(di, urgency);
  52
  53         uctd_get_best_child(descent);
  54 }