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.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)
  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                         /* Play pass only if we can afford scoring */
  26                         if (is_pass(ni->coord) && !uct_pass_is_safe(p->uct, b, color, p->uct->pass_all_alive))
  27                                 continue;
  28                         nbest = ni;
  29                 }
  30         return nbest;
  31 }
  32
  33 /* Return the node with best value instead of best explored. We must use the heuristic
  34  * value (using prior and possibly rave), because the raw value is meaningless for
  35  * nodes evaluated rarely.
  36  * This function is called while the tree is updated by other threads */
  37 struct tree_node *
  38 uctp_generic_winner(struct uct_policy *p, struct tree *tree, struct tree_node *node)
  39 {
  40         if (!p->evaluate)
  41                 return NULL;
  42         bool allow_pass = false; /* At worst forces some extra playouts at the end */
  43         void *state; /* TODO: remove this unused parameter. */
  44         int parity = ((node->depth ^ tree->root->depth) & 1) ? -1 : 1;
  45
  46         uctd_try_node_children(node, allow_pass, ni, urgency) {
  47                 urgency = p->evaluate(p, state, tree, ni, parity);
  48         } uctd_set_best_child(ni, urgency);
  49
  50         return uctd_get_best_child();
  51 }