From 0047f19e0a2332a30d4d0b214fb2d07dcea71dfe Mon Sep 17 00:00:00 2001
From: Petr Baudis <pasky@ucw.cz>
Date: Fri, 4 Apr 2008 03:33:40 +0200
Subject: [PATCH] UCT: Keep prior moves in separate stats

This e.g. keeps grandfather from picking up prior noise.
---
 uct/policy/ucb1.c     | 20 ++++++++++----------
 uct/policy/ucb1amaf.c |  2 +-
 uct/tree.h            |  1 +
 3 files changed, 12 insertions(+), 11 deletions(-)

diff --git a/uct/policy/ucb1.c b/uct/policy/ucb1.c
index da66c27..28b4fe4 100644
--- a/uct/policy/ucb1.c
+++ b/uct/policy/ucb1.c
@@ -81,8 +81,8 @@ ucb1_prior(struct uct_policy *p, struct tree *tree, struct tree_node *node, stru
 #if 0
 	/* Q_{even} */
 	/* This somehow does not work at all. */
-	node->u.playouts += p->eqex;
-	node->u.wins += p->eqex / 2;
+	node->prior.playouts += p->eqex;
+	node->prior.wins += p->eqex / 2;
 #endif
 
 	/* Q_{grandparent} */
@@ -91,8 +91,8 @@ ucb1_prior(struct uct_policy *p, struct tree *tree, struct tree_node *node, stru
 		for (struct tree_node *ni = gpp->children; ni; ni = ni->sibling) {
 			/* Be careful not to emphasize too random results. */
 			if (ni->coord == node->coord && ni->u.playouts > pp->eqex) {
-				node->u.playouts += pp->eqex;
-				node->u.wins += pp->eqex * ni->u.wins / ni->u.playouts;
+				node->prior.playouts += pp->eqex;
+				node->prior.wins += pp->eqex * ni->u.wins / ni->u.playouts;
 				node->hints |= 1;
 			}
 		}
@@ -108,15 +108,15 @@ ucb1_prior(struct uct_policy *p, struct tree *tree, struct tree_node *node, stru
 	if (!isnan(assess)) {
 		if (parity < 0)
 			assess = 1 - assess;
-		node->u.playouts += pp->eqex;
-		node->u.wins += pp->eqex * assess;
+		node->prior.playouts += pp->eqex;
+		node->prior.wins += pp->eqex * assess;
 		node->hints |= 2;
 	}
 
-	if (node->u.playouts)
-		node->u.value = (float) node->u.wins / node->u.playouts;
+	if (node->prior.playouts)
+		node->prior.value = (float) node->prior.wins / node->prior.playouts;
 
-	//fprintf(stderr, "%s,%s prior: %d/%d = %f (%f)\n", coord2sstr(node->parent->coord, b), coord2sstr(node->coord, b), node->u.wins, node->u.playouts, node->u.value, assess);
+	//fprintf(stderr, "%s,%s prior: %d/%d = %f (%f)\n", coord2sstr(node->parent->coord, b), coord2sstr(node->coord, b), node->prior.wins, node->prior.playouts, node->prior.value, assess);
 }
 
 void
@@ -129,7 +129,7 @@ ucb1_update(struct uct_policy *p, struct tree_node *node, enum stone color, stru
 	for (; node; node = node->parent) {
 		node->u.playouts++;
 		node->u.wins += result;
-		node->u.value = (float)node->u.wins / node->u.playouts;
+		node->u.value = ((float)node->u.wins + (float)node->prior.wins) / (node->u.playouts + node->prior.playouts);
 	}
 }
 
diff --git a/uct/policy/ucb1amaf.c b/uct/policy/ucb1amaf.c
index 2443824..eef4fdd 100644
--- a/uct/policy/ucb1amaf.c
+++ b/uct/policy/ucb1amaf.c
@@ -39,7 +39,7 @@ update_node(struct uct_policy *p, struct tree_node *node, int result)
 {
 	node->u.playouts++;
 	node->u.wins += result;
-	node->u.value = (float)node->u.wins / node->u.playouts;
+	node->u.value = ((float)node->u.wins + (float)node->prior.wins) / (node->u.playouts + node->prior.playouts);
 }
 
 void
diff --git a/uct/tree.h b/uct/tree.h
index 2d9eced..d1b4238 100644
--- a/uct/tree.h
+++ b/uct/tree.h
@@ -35,6 +35,7 @@ struct tree_node {
 	coord_t coord;
 
 	struct move_stats u;
+	struct move_stats prior;
 	int hints;
 };
 
-- 
2.11.4.GIT