Merge pull request #661 from karlnapf/master

fixes to make MKL working with X-validation
shogun-toolbox · Jul 20, 2012 · 823ef52 · 823ef52
2 parents d5d5151 + cea8bcd
commit 823ef52
Show file tree

Hide file tree

Showing 6 changed files with 82 additions and 18 deletions.
diff --git a/src/shogun/evaluation/CrossValidation.cpp b/src/shogun/evaluation/CrossValidation.cpp
@@ -63,6 +63,15 @@ CEvaluationResult* CCrossValidation::evaluate(CModelSelectionOutput* ms_output)
 {
 	SG_DEBUG("entering %s::evaluate()\n", get_name());
 
+	REQUIRE(m_machine, "%s::evaluate() is only possible if a machine is "
+			"attached\n", get_name());
+
+	REQUIRE(m_features, "%s::evaluate() is only possible if features are "
+			"attached\n", get_name());
+
+	REQUIRE(m_labels, "%s::evaluate() is only possible if labels are "
+			"attached\n", get_name());
+
 	/* if for some reason the do_unlock_frag is set, unlock */
 	if (m_do_unlock)
 	{
@@ -250,7 +259,9 @@ float64_t CCrossValidation::evaluate_one_run(CModelSelectionOutput* ms_output)
 			}
 
 			/* train machine on training features and remove subset */
+			SG_DEBUG("starting training\n");
 			m_machine->train(m_features);
+			SG_DEBUG("finished training\n");
 			if (ms_output)
 			{
 				ms_output->output_train_indices(inverse_subset_indices);
@@ -276,7 +287,10 @@ float64_t CCrossValidation::evaluate_one_run(CModelSelectionOutput* ms_output)
 			}
 
 			/* apply machine to test features and remove subset */
+			SG_DEBUG("starting evaluation\n");
+			SG_DEBUG("%p\n", m_features);
 			CLabels* result_labels=m_machine->apply(m_features);
+			SG_DEBUG("finished evaluation\n");
 			m_features->remove_subset();
 			SG_REF(result_labels);
 

diff --git a/src/shogun/labels/Labels.cpp b/src/shogun/labels/Labels.cpp
@@ -6,7 +6,7 @@
  *
  * Written (W) 1999-2009 Soeren Sonnenburg
  * Written (W) 1999-2008 Gunnar Raetsch
- * Written (W) 2011 Heiko Strathmann
+ * Written (W) 2011-2012 Heiko Strathmann
  * Copyright (C) 1999-2009 Fraunhofer Institute FIRST and Max-Planck-Society
  */
 
@@ -54,8 +54,8 @@ void CLabels::remove_all_subsets()
 
 float64_t CLabels::get_confidence(int32_t idx)
 {
-	int32_t real_num=m_subset_stack->subset_idx_conversion(idx);
 	ASSERT(m_confidences.vector && idx<get_num_labels());
+	int32_t real_num=m_subset_stack->subset_idx_conversion(idx);
 	return m_confidences.vector[real_num];
 }
 

diff --git a/src/shogun/machine/KernelMachine.cpp b/src/shogun/machine/KernelMachine.cpp
@@ -250,7 +250,10 @@ CBinaryLabels* CKernelMachine::apply_binary(CFeatures* data)
 
 SGVector<float64_t> CKernelMachine::apply_get_outputs(CFeatures* data)
 {
-	REQUIRE(kernel, "No kernel assigned!\n");
+	SG_DEBUG("entering %s::apply_get_outputs(%s at %p)\n",
+			get_name(), data ? data->get_name() : "NULL", data);
+
+	REQUIRE(kernel, "%s::apply_get_outputs(): No kernel assigned!\n");
 
 	if (!kernel->get_num_vec_lhs())
 	{
@@ -263,12 +266,22 @@ SGVector<float64_t> CKernelMachine::apply_get_outputs(CFeatures* data)
 	if (data)
 	{
 		CFeatures* lhs=kernel->get_lhs();
-		REQUIRE(lhs, "%s: No left hand side specified\n", get_name());
+		REQUIRE(lhs, "%s::apply_get_outputs(): No left hand side specified\n",
+				get_name());
 		kernel->init(lhs, data);
 		SG_UNREF(lhs);
 	}
 
-	int32_t num_vectors=kernel->get_num_vec_rhs();
+	/* using the features to get num vectors is far safer than using the kernel
+	 * since SHOGUNs kernel num_rhs/num_lhs is buggy (CombinedKernel for ex.)
+	 * Might be worth investigating why
+	 * kernel->get_num_rhs() != rhs->get_num_vectors()
+	 * However, the below version works
+	 * TODO Heiko Strathmann
+	 */
+	CFeatures* rhs=kernel->get_rhs();
+	int32_t num_vectors=rhs->get_num_vectors();
+	SG_UNREF(rhs)
 
 	SGVector<float64_t> output(num_vectors);
 
@@ -380,6 +393,9 @@ SGVector<float64_t> CKernelMachine::apply_get_outputs(CFeatures* data)
 			SG_DONE();
 	}
 
+	SG_DEBUG("leaving %s::apply_get_outputs(%s at %p)\n",
+			get_name(), data ? data->get_name() : "NULL", data);
+
 	return output;
 }
 

diff --git a/src/shogun/machine/KernelMulticlassMachine.h b/src/shogun/machine/KernelMulticlassMachine.h
@@ -5,6 +5,7 @@
  * (at your option) any later version.
  *
  * Written (W) 1999-2012 Soeren Sonnenburg and Sergey Lisitsyn
+ * Written (W) 2012 Heiko Strathmann
  * Copyright (C) 1999-2012 Fraunhofer Institute FIRST and Max-Planck-Society
  */
 
@@ -104,14 +105,22 @@ class CKernelMulticlassMachine : public CMulticlassMachine
 		virtual bool init_machines_for_apply(CFeatures* data)
 		{
 			if (data)
-				m_kernel->init(m_kernel->get_lhs(),data);
+			{
+				/* set data to rhs for this kernel */
+				CFeatures* lhs=m_kernel->get_lhs();
+				m_kernel->init(lhs, data);
+				SG_UNREF(lhs);
+			}
 
+			/* set kernel to all sub-machines */
 			for (int32_t i=0; i<m_machines->get_num_elements(); i++)
 			{
-				CKernelMachine *machine = (CKernelMachine*)m_machines->get_element(i);
+				CKernelMachine *machine=
+						(CKernelMachine*)m_machines->get_element(i);
 				machine->set_kernel(m_kernel);
 				SG_UNREF(machine);
 			}
+
 			return true;
 		}
 

diff --git a/src/shogun/machine/Machine.cpp b/src/shogun/machine/Machine.cpp
@@ -160,19 +160,34 @@ void CMachine::data_unlock()
 
 CLabels* CMachine::apply(CFeatures* data)
 {
+	SG_DEBUG("entering %s::apply(%s at %p)\n",
+			get_name(), data ? data->get_name() : "NULL", data);
+
+	CLabels* result=NULL;
+
 	switch (get_machine_problem_type())
 	{
 		case PT_BINARY:
-			return apply_binary(data);
+			result=apply_binary(data);
+			break;
 		case PT_REGRESSION:
-			return apply_regression(data);
+			result=apply_regression(data);
+			break;
 		case PT_MULTICLASS:
-			return apply_multiclass(data);
+			result=apply_multiclass(data);
+			break;
 		case PT_STRUCTURED:
-			return apply_structured(data);
-		default: SG_ERROR("Unknown problem type");
+			result=apply_structured(data);
+			break;
+		default:
+			SG_ERROR("Unknown problem type");
+			break;
 	}
-	return NULL;
+
+	SG_DEBUG("leaving %s::apply(%s at %p)\n",
+			get_name(), data ? data->get_name() : "NULL", data);
+
+	return result;
 }
 
 CLabels* CMachine::apply_locked(SGVector<index_t> indices)

diff --git a/src/shogun/machine/MulticlassMachine.cpp b/src/shogun/machine/MulticlassMachine.cpp
@@ -91,14 +91,22 @@ float64_t CMulticlassMachine::get_submachine_output(int32_t i, int32_t num)
 
 CMulticlassLabels* CMulticlassMachine::apply_multiclass(CFeatures* data)
 {
+	SG_DEBUG("entering %s::apply_multiclass(%s at %p)\n",
+			get_name(), data ? data->get_name() : "NULL", data);
+
+	CMulticlassLabels* return_labels=NULL;
+
 	if (data)
 		init_machines_for_apply(data);
 	else
 		init_machines_for_apply(NULL);
 
 	if (is_ready())
 	{
-		int32_t num_vectors=get_num_rhs_vectors();
+		/* num vectors depends on whether data is provided */
+		int32_t num_vectors=data ? data->get_num_vectors() :
+				get_num_rhs_vectors();
+
 		int32_t num_machines=m_machines->get_num_elements();
 		if (num_machines <= 0)
 			SG_ERROR("num_machines = %d, did you train your machine?", num_machines);
@@ -123,13 +131,15 @@ CMulticlassLabels* CMulticlassMachine::apply_multiclass(CFeatures* data)
 
 		SG_FREE(outputs);
 
-		return result;
+		return_labels=result;
 	}
 	else
-	{
 		SG_ERROR("Not ready");
-		return NULL;
-	}
+
+
+	SG_DEBUG("leaving %s::apply_multiclass(%s at %p)\n",
+				get_name(), data ? data->get_name() : "NULL", data);
+	return return_labels;
 }
 
 bool CMulticlassMachine::train_machine(CFeatures* data)