Fixed a minor bug in how the cross validation accuracy was being

computed.

dlib/svm/cross_validate_assignment_trainer.h

@@ -114,13 +114,14 @@ namespace dlib
         const long num_in_test  = samples.size()/folds;
         const long num_in_train = samples.size() - num_in_test;
 
-        running_stats<double> rs;
 
         std::vector<sample_type> samples_test, samples_train;
         std::vector<label_type> labels_test, labels_train;
 
 
         long next_test_idx = 0;
+        double total_right = 0;
+        double total = 0;
 
 
         for (long i = 0; i < folds; ++i)
@@ -148,13 +149,28 @@ namespace dlib
             }
 
 
-            rs.add(test_assignment_function(trainer.train(samples_train,labels_train),
-                                            samples_test,
-                                            labels_test));
+            const typename trainer_type::trained_function_type& df = trainer.train(samples_train,labels_train);
+
+            // check how good df is on the test data
+            for (unsigned long i = 0; i < samples_test.size(); ++i)
+            {
+                const std::vector<long>& out = df(samples_test[i]);
+                for (unsigned long j = 0; j < out.size(); ++j)
+                {
+                    if (out[j] == labels_test[i][j])
+                        ++total_right;
+
+                    ++total;
+                }
+            }
 
         } // for (long i = 0; i < folds; ++i)
 
-        return rs.mean();
+        if (total != 0)
+            return total_right/total;
+        else
+            return 1;
+
     }
 
 // ----------------------------------------------------------------------------------------