Merge pull request #658 from karlnapf/master

cross-validation for combined features aka MKL
shogun-toolbox · Jul 18, 2012 · 01d6292 · 01d6292
2 parents 46b4d2c + 4f86bd8
commit 01d6292
Show file tree

Hide file tree

Showing 14 changed files with 221 additions and 20 deletions.
diff --git a/src/interfaces/modular/Features.i b/src/interfaces/modular/Features.i
@@ -20,6 +20,7 @@
 /* These functions return new Objects */
 %newobject get_transposed();
 %newobject create_merged_copy(CFeatures* other);
+%newobject copy_subset(SGVector<index_t> indices);
 
 #ifdef USE_SWIG_DIRECTORS
 %feature("director") shogun::CDirectorDotFeatures;

diff --git a/src/shogun/classifier/mkl/MKL.cpp b/src/shogun/classifier/mkl/MKL.cpp
@@ -196,7 +196,11 @@ bool CMKL::train_machine(CFeatures* data)
 	if (data)
 	{
 		if (m_labels->get_num_labels() != data->get_num_vectors())
-			SG_ERROR("Number of training vectors does not match number of labels\n");
+		{
+			SG_ERROR("%s::train_machine(): Number of training vectors (%d) does"
+					" not match number of labels (%d)\n", get_name(),
+					data->get_num_vectors(), m_labels->get_num_labels());
+		}
 		kernel->init(data, data);
 	}
 

diff --git a/src/shogun/classifier/mkl/MKLMulticlass.cpp b/src/shogun/classifier/mkl/MKLMulticlass.cpp
@@ -329,8 +329,11 @@ bool CMKLMulticlass::train_machine(CFeatures* data)
 	if (data)
 	{
 		if (m_labels->get_num_labels() != data->get_num_vectors())
-			SG_ERROR("Number of training vectors does not match number of "
-					"labels\n");
+		{
+			SG_ERROR("%s::train_machine(): Number of training vectors (%d) does"
+					" not match number of labels (%d)\n", get_name(),
+					data->get_num_vectors(), m_labels->get_num_labels());
+		}
 		m_kernel->init(data, data);
 	}
 

diff --git a/src/shogun/classifier/svm/CPLEXSVM.cpp b/src/shogun/classifier/svm/CPLEXSVM.cpp
@@ -39,7 +39,11 @@ bool CCPLEXSVM::train_machine(CFeatures* data)
 	if (data)
 	{
 		if (m_labels->get_num_labels() != data->get_num_vectors())
-			SG_ERROR("Number of training vectors does not match number of labels\n");
+		{
+			SG_ERROR("%s::train_machine(): Number of training vectors (%d) does"
+					" not match number of labels (%d)\n", get_name(),
+					data->get_num_vectors(), m_labels->get_num_labels());
+		}
 		kernel->init(data, data);
 	}
 

diff --git a/src/shogun/classifier/svm/LibSVM.cpp b/src/shogun/classifier/svm/LibSVM.cpp
@@ -40,7 +40,11 @@ bool CLibSVM::train_machine(CFeatures* data)
 	if (data)
 	{
 		if (m_labels->get_num_labels() != data->get_num_vectors())
-			SG_ERROR("Number of training vectors does not match number of labels\n");
+		{
+			SG_ERROR("%s::train_machine(): Number of training vectors (%d) does"
+					" not match number of labels (%d)\n", get_name(),
+					data->get_num_vectors(), m_labels->get_num_labels());
+		}
 		kernel->init(data, data);
 	}
 

diff --git a/src/shogun/classifier/svm/SVMLight.cpp b/src/shogun/classifier/svm/SVMLight.cpp
@@ -208,7 +208,11 @@ bool CSVMLight::train_machine(CFeatures* data)
 	if (data)
 	{
 		if (m_labels->get_num_labels() != data->get_num_vectors())
-			SG_ERROR("Number of training vectors does not match number of labels\n");
+		{
+			SG_ERROR("%s::train_machine(): Number of training vectors (%d) does"
+					" not match number of labels (%d)\n", get_name(),
+					data->get_num_vectors(), m_labels->get_num_labels());
+		}
 		kernel->init(data, data);
 	}
 

diff --git a/src/shogun/features/CombinedFeatures.cpp b/src/shogun/features/CombinedFeatures.cpp
@@ -6,11 +6,14 @@
  *
  * Written (W) 1999-2009 Soeren Sonnenburg
  * Written (W) 1999-2008 Gunnar Raetsch
+ * Written (W) 2012 Heiko Strathmann
  * Copyright (C) 1999-2009 Fraunhofer Institute FIRST and Max-Planck-Society
  */
 
 #include <shogun/features/CombinedFeatures.h>
 #include <shogun/io/SGIO.h>
+#include <shogun/lib/Set.h>
+#include <shogun/lib/Map.h>
 
 using namespace shogun;
 
@@ -153,8 +156,11 @@ bool CCombinedFeatures::insert_feature_obj(CFeatures* obj)
 	ASSERT(obj);
 	int32_t n=obj->get_num_vectors();
 
-	if (num_vec>0 && n!=num_vec)
-		SG_ERROR("Number of feature vectors does not match (expected %d, obj has %d)\n", num_vec, n);
+	if (get_num_vectors()>0 && n!=get_num_vectors())
+	{
+		SG_ERROR("Number of feature vectors does not match (expected %d, "
+				"obj has %d)\n", get_num_vectors(), n);
+	}
 
 	num_vec=n;
 	return feature_list->insert_element(obj);
@@ -165,8 +171,11 @@ bool CCombinedFeatures::append_feature_obj(CFeatures* obj)
 	ASSERT(obj);
 	int32_t n=obj->get_num_vectors();
 
-	if (num_vec>0 && n!=num_vec)
-		SG_ERROR("Number of feature vectors does not match (expected %d, obj has %d)\n", num_vec, n);
+	if (get_num_vectors()>0 && n!=get_num_vectors())
+	{
+		SG_ERROR("Number of feature vectors does not match (expected %d, "
+				"obj has %d)\n", get_num_vectors(), n);
+	}
 
 	num_vec=n;
 	return feature_list->append_element(obj);
@@ -199,6 +208,10 @@ void CCombinedFeatures::init()
 
 CFeatures* CCombinedFeatures::create_merged_copy(CFeatures* other)
 {
+	/* TODO, if all features are the same, only one copy should be created
+	 * in memory */
+	SG_WARNING("Heiko Strathmann: FIXME, unefficient!\n");
+
 	SG_DEBUG("entering %s::create_merged_copy()\n", get_name());
 	if (get_feature_type()!=other->get_feature_type() ||
 			get_feature_class()!=other->get_feature_class() ||
@@ -238,3 +251,121 @@ CFeatures* CCombinedFeatures::create_merged_copy(CFeatures* other)
 	SG_DEBUG("leaving %s::create_merged_copy()\n", get_name());
 	return result;
 }
+
+void CCombinedFeatures::add_subset(SGVector<index_t> subset)
+{
+	SG_DEBUG("entering %s::add_subset()\n", get_name());
+	CSet<CFeatures*>* processed=new CSet<CFeatures*>();
+
+	CFeatures* current=get_first_feature_obj();
+	while (current)
+	{
+		if (!processed->contains(current))
+		{
+			/* remember that subset was added here */
+			current->add_subset(subset);
+			processed->add(current);
+			SG_DEBUG("adding subset to %s at %p\n",
+					current->get_name(), current);
+		}
+		SG_UNREF(current);
+		current=get_next_feature_obj();
+	}
+
+	/* also add subset to local stack to have it for easy access */
+	m_subset_stack->add_subset(subset);
+
+	subset_changed_post();
+	SG_UNREF(processed);
+	SG_DEBUG("leaving %s::add_subset()\n", get_name());
+}
+
+void CCombinedFeatures::remove_subset()
+{
+	SG_DEBUG("entering %s::remove_subset()\n", get_name());
+	CSet<CFeatures*>* processed=new CSet<CFeatures*>();
+
+	CFeatures* current=get_first_feature_obj();
+	while (current)
+	{
+		if (!processed->contains(current))
+		{
+			/* remember that subset was added here */
+			current->remove_subset();
+			processed->add(current);
+			SG_DEBUG("removing subset from %s at %p\n",
+					current->get_name(), current);
+		}
+		SG_UNREF(current);
+		current=get_next_feature_obj();
+	}
+
+	/* also remove subset from local stack to have it for easy access */
+	m_subset_stack->remove_subset();
+
+	subset_changed_post();
+	SG_UNREF(processed);
+	SG_DEBUG("leaving %s::remove_subset()\n", get_name());
+}
+
+void CCombinedFeatures::remove_all_subsets()
+{
+	SG_DEBUG("entering %s::remove_all_subsets()\n", get_name());
+	CSet<CFeatures*>* processed=new CSet<CFeatures*>();
+
+	CFeatures* current=get_first_feature_obj();
+	while (current)
+	{
+		if (!processed->contains(current))
+		{
+			/* remember that subset was added here */
+			current->remove_all_subsets();
+			processed->add(current);
+			SG_DEBUG("removing all subsets from %s at %p\n",
+					current->get_name(), current);
+		}
+		SG_UNREF(current);
+		current=get_next_feature_obj();
+	}
+
+	/* also remove subsets from local stack to have it for easy access */
+	m_subset_stack->remove_all_subsets();
+
+	subset_changed_post();
+	SG_UNREF(processed);
+	SG_DEBUG("leaving %s::remove_all_subsets()\n", get_name());
+}
+
+CFeatures* CCombinedFeatures::copy_subset(SGVector<index_t> indices)
+{
+	/* this is returned with the results of copy_subset of sub-features */
+	CCombinedFeatures* result=new CCombinedFeatures();
+
+	/* map to only copy same feature objects once */
+	CMap<CFeatures*, CFeatures*>* processed=new CMap<CFeatures*, CFeatures*>();
+	CFeatures* current=get_first_feature_obj();
+	while (current)
+	{
+		CFeatures* new_element=NULL;
+
+		/* only copy if not done yet, otherwise, use old copy */
+		if (!processed->contains(current))
+		{
+			new_element=current->copy_subset(indices);
+			processed->add(current, new_element);
+		}
+		else
+			new_element=processed->get_element(current);
+
+		/* add to result */
+		result->append_feature_obj(new_element);
+
+		SG_UNREF(current);
+		current=get_next_feature_obj();
+	}
+
+	SG_UNREF(processed);
+
+	SG_REF(result);
+	return result;
+}
diff --git a/src/shogun/features/CombinedFeatures.h b/src/shogun/features/CombinedFeatures.h
@@ -6,6 +6,7 @@
  *
  * Written (W) 1999-2009 Soeren Sonnenburg
  * Written (W) 1999-2008 Gunnar Raetsch
+ * Written (W) 2012 Heiko Strathmann
  * Copyright (C) 1999-2009 Fraunhofer Institute FIRST and Max-Planck-Society
  */
 
@@ -27,6 +28,11 @@ class CListElement;
  * It keeps pointers to the added sub-features and is especially useful to
  * combine kernels working on different domains (c.f. CCombinedKernel) and to
  * combine kernels looking at independent features.
+ *
+ * Subsets are supported: All actions will just be given through to all
+ * sub-features. Only once per sub-feature instance, i.e. if there are two
+ * sub-features that are the same instance, the subset action will only be
+ * performed once.
  */
 class CCombinedFeatures : public CFeatures
 {
@@ -69,7 +75,8 @@ class CCombinedFeatures : public CFeatures
 		 */
 		inline virtual int32_t get_num_vectors() const
 		{
-			return num_vec;
+			return m_subset_stack->has_subsets()
+					? m_subset_stack->get_size() : num_vec;
 		}
 
 		/** get memory footprint of one feature
@@ -158,6 +165,38 @@ class CCombinedFeatures : public CFeatures
 		 */
 		CFeatures* create_merged_copy(CFeatures* other);
 
+		/** adds a subset of indices on top of the current subsets (possibly
+		 * subset o subset. Calls subset_changed_post() afterwards.
+		 * Adds the subset to all sub-features
+		 *
+		 * @param subset subset of indices to add
+		 * */
+		virtual void add_subset(SGVector<index_t> subset);
+
+		/** removes that last added subset from subset stack, if existing
+		 * Calls subset_changed_post() afterwards
+		 *
+		 * Removes the subset from all sub-features
+		 * */
+		virtual void remove_subset();
+
+		/** removes all subsets
+		 * Calls subset_changed_post() afterwards
+		 *
+		 * Removes all subsets of all sub-features
+		 * */
+		virtual void remove_all_subsets();
+
+		/** Creates a new CFeatures instance containing copies of the elements
+		 * which are specified by the provided indices.
+		 * Simply creates a combined features instance where all sub-features
+		 * are the results of their copy_subset calls
+		 *
+		 * @param indices indices of feature elements to copy
+		 * @return new CFeatures instance with copies of feature data
+		 */
+		virtual CFeatures* copy_subset(SGVector<index_t> indices);
+
 		/** @return object name */
 		inline virtual const char* get_name() const { return "CombinedFeatures"; }
 

diff --git a/src/shogun/features/DenseFeatures.cpp b/src/shogun/features/DenseFeatures.cpp
@@ -557,7 +557,9 @@ template<class ST> CFeatures* CDenseFeatures<ST>::copy_subset(SGVector<index_t>
 				num_features*sizeof(ST));
 	}
 
-	return new CDenseFeatures(feature_matrix_copy);
+	CFeatures* result=new CDenseFeatures(feature_matrix_copy);
+	SG_REF(result);
+	return result;
 }
 
 template<class ST> ST* CDenseFeatures<ST>::compute_feature_vector(int32_t num, int32_t& len,

diff --git a/src/shogun/features/Features.cpp b/src/shogun/features/Features.cpp
@@ -368,8 +368,8 @@ void CFeatures::remove_all_subsets()
 
 CFeatures* CFeatures::copy_subset(SGVector<index_t> indices)
 {
-	SG_ERROR("copy_subset and therefore model storage of CMachine "
-			"(required for cross-validation and model-selection is ",
-			"not yet implemented for feature type %s\n", get_name());
+	SG_ERROR("%s::copy_subset(): copy_subset and therefore model storage of "
+			"CMachine (required for cross-validation and model-selection is "
+			"not yet implemented yet. Ask developers!\n", get_name());
 	return NULL;
 }
diff --git a/src/shogun/features/SparseFeatures.cpp b/src/shogun/features/SparseFeatures.cpp
@@ -1070,7 +1070,9 @@ template<class ST> CFeatures* CSparseFeatures<ST>::copy_subset(SGVector<index_t>
 		free_sparse_feature_vector(index);
 	}
 
-	return new CSparseFeatures<ST>(matrix_copy);
+	CFeatures* result=new CSparseFeatures<ST>(matrix_copy);
+	SG_REF(result);
+	return result;
 }
 
 template<class ST> SGSparseVectorEntry<ST>* CSparseFeatures<ST>::compute_sparse_feature_vector(int32_t num,

diff --git a/src/shogun/features/StringFeatures.cpp b/src/shogun/features/StringFeatures.cpp
@@ -1611,6 +1611,8 @@ template<class ST> CFeatures* CStringFeatures<ST>::copy_subset(SGVector<index_t>
 	/* max string length may have changed */
 	result->determine_maximum_string_length();
 
+	SG_REF(result);
+
 	return result;
 }
 

diff --git a/src/shogun/features/SubsetStack.cpp b/src/shogun/features/SubsetStack.cpp
@@ -54,15 +54,19 @@ void CSubsetStack::add_subset(SGVector<index_t> subset)
 		CSubset* latest=(CSubset*)m_active_subsets_stack->get_last_element();
 		if (subset.vlen>latest->m_subset_idx.vlen)
 		{
-			SG_ERROR("Error in %s::add_subset(): Provided index vector is "
+			subset.display_vector("subset");
+			latest->m_subset_idx.display_vector("last on stack");
+			SG_ERROR("%s::add_subset(): Provided index vector is "
 					"larger than the subsets on the stubset stack!\n", get_name());
 		}
 
 		/* check for range of indices */
 		index_t max_index=SGVector<index_t>::max(subset.vector, subset.vlen);
 		if (max_index>=latest->m_subset_idx.vlen)
 		{
-			SG_ERROR("Error in %s::add_subset(): Provided index vector contains"
+			subset.display_vector("subset");
+			latest->m_subset_idx.display_vector("last on stack");
+			SG_ERROR("%s::add_subset(): Provided index vector contains"
 					" indices larger than possible range!\n", get_name());
 		}
 

diff --git a/src/shogun/multiclass/GMNPSVM.cpp b/src/shogun/multiclass/GMNPSVM.cpp
@@ -58,9 +58,10 @@ bool CGMNPSVM::train_machine(CFeatures* data)
 
 	if (data)
 	{
-		if (data->get_num_vectors() != m_labels->get_num_labels())
+		if (m_labels->get_num_labels() != data->get_num_vectors())
 		{
-			SG_ERROR("Numbert of vectors (%d) does not match number of labels (%d)\n",
+			SG_ERROR("%s::train_machine(): Number of training vectors (%d) does"
+					" not match number of labels (%d)\n", get_name(),
 					data->get_num_vectors(), m_labels->get_num_labels());
 		}
 		m_kernel->init(data, data);