X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=bayeswf.cpp;h=cc591832a556f023ea6aa364a44ce5422a0b0740;hb=1fac0677bbcba5dec26e11aa661c97e38f6c3d40;hp=ca205ef47e4eb9cf75f9b27370d280e60d431b46;hpb=0e3311bf0601a8db66058439bd1fff2a9cac2e58;p=wloh

diff --git a/bayeswf.cpp b/bayeswf.cpp
index ca205ef..cc59183 100644
--- a/bayeswf.cpp
+++ b/bayeswf.cpp
@@ -13,7 +13,7 @@
 using namespace std;
 using namespace Eigen;
 
-#define PRIOR_MU 1500
+#define PRIOR_MU 500
 #define PRIOR_WEIGHT 1.0
 #define MAX_PLAYERS 4096
 #define DUMP_RAW 0
@@ -199,13 +199,14 @@ float compute_total_logl(float *mu, int num_players)
  *
  * Note that this does not depend on mu or the margin at all.
  */
-double hessian[MAX_PLAYERS * MAX_PLAYERS];
+Matrix<float, Dynamic, Dynamic> hessian;
 void construct_hessian(const float *mu, int num_players)
 {
-	memset(hessian, 0, sizeof(hessian));
+	hessian = Matrix<float, Dynamic, Dynamic>(num_players, num_players);
+	hessian.fill(0.0f);
 
 	for (int i = 0; i < num_players; ++i) {
-		hessian[i * num_players + i] += 1.0f / (prior_sigma * prior_sigma);
+		hessian(i, i) += 1.0f / (prior_sigma * prior_sigma);
 	}
 	for (unsigned i = 0; i < all_matches.size(); ++i) {
 		const match &m = all_matches[i];
@@ -216,30 +217,32 @@ void construct_hessian(const float *mu, int num_players)
 		double sigma_sq = global_sigma * global_sigma;
 		float w = m.weight;
 
-		hessian[p1 * num_players + p2] -= w / sigma_sq;
-		hessian[p2 * num_players + p1] -= w / sigma_sq;
+		hessian(p1, p2) -= w / sigma_sq;
+		hessian(p2, p1) -= w / sigma_sq;
 
-		hessian[p1 * num_players + p1] += w / sigma_sq;
-		hessian[p2 * num_players + p2] += w / sigma_sq;
+		hessian(p1, p1) += w / sigma_sq;
+		hessian(p2, p2) += w / sigma_sq;
 	}
 }
 
 // Compute uncertainty (stddev) of mu estimates, which is sqrt((H^-1)_ii),
 // where H is the Hessian (see construct_hessian()).
-void compute_mu_uncertainty(const float *mu, int num_players)
+void compute_mu_uncertainty(const float *mu, const vector<string> &players)
 {
-	Matrix<float, Dynamic, Dynamic> h(num_players, num_players);
-	for (int i = 0; i < num_players; ++i) {
-		for (int j = 0; j < num_players; ++j) {
-			h(i, j) = hessian[i * num_players + j];
-		}
-	}
-
 	// FIXME: Use pseudoinverse if applicable.
-	Matrix<float, Dynamic, Dynamic> ih = h.inverse();
-	for (int i = 0; i < num_players; ++i) {
+	Matrix<float, Dynamic, Dynamic> ih = hessian.inverse();
+	for (unsigned i = 0; i < players.size(); ++i) {
 		mu_stddev[i] = sqrt(ih(i, i));
 	}
+
+	for (unsigned i = 0; i < players.size(); ++i) {
+		for (unsigned j = 0; j < players.size(); ++j) {
+			printf("covariance %s %s %f\n",
+			       players[i].c_str(),
+			       players[j].c_str(),
+			       ih(i, j));
+		}
+	}
 }
 
 int main(int argc, char **argv)
@@ -337,7 +340,7 @@ int main(int argc, char **argv)
 		sumdiff += (global_sigma - old_global_sigma) * (global_sigma - old_global_sigma);
 		if (sumdiff < EPSILON) {
 			//fprintf(stderr, "Converged after %d iterations. Stopping.\n", j);
-			printf("%d 0 -1\n", j + 1);
+			printf("aux_param num_iterations %d\n", j + 1);
 			break;
 		}
 	}
@@ -346,13 +349,13 @@ int main(int argc, char **argv)
 	dump_raw(mu, num_players);
 #else
 	construct_hessian(mu, num_players);
-	compute_mu_uncertainty(mu, num_players);
+	compute_mu_uncertainty(mu, players);
 	dump_scores(players, mu, mu_stddev, num_players);
 	//fprintf(stderr, "Optimal sigma: %f (two-player: %f)\n", sigma[0], sigma[0] * sqrt(2.0f));
-	printf("%f 0 -2\n", global_sigma / sqrt(2.0f));
-	printf("%f 0 -3\n", prior_sigma);
+	printf("aux_param score_stddev %f\n", global_sigma / sqrt(2.0f));
+	printf("aux_param rating_prior_stddev %f\n", prior_sigma);
 
 	float total_logl = compute_total_logl(mu, num_players);
-	printf("%f 0 -4\n", total_logl);
+	printf("aux_param total_log_likelihood %f\n", total_logl);
 #endif
 }