Code review: change "validation" to "test"

shogun-toolbox · Apr 11, 2018 · 0a2a42a · 0a2a42a
1 parent 26cb0a1
commit 0a2a42a
Show file tree

Hide file tree

Showing 12 changed files with 21 additions and 24 deletions.
diff --git a/doc/ipython-notebooks/evaluation/xval_modelselection.ipynb b/doc/ipython-notebooks/evaluation/xval_modelselection.ipynb
@@ -197,7 +197,7 @@
     "    y=[]\n",
     "    lab=[]\n",
     "    for j in range(num):\n",
-    "        indices=split.validation(j)\n",
+    "        indices=split.test(j)\n",
     "        x_=[]\n",
     "        y_=[]\n",
     "        lab_=[]\n",

diff --git a/doc/ipython-notebooks/multiclass/KNN.ipynb b/doc/ipython-notebooks/multiclass/KNN.ipynb
@@ -289,7 +289,7 @@
     "    time_test = np.zeros(accuracy.shape)\n",
     "    for i in range(Nsplit):\n",
     "        idx_train = split.train(i)\n",
-    "        idx_test  = split.validation(i)\n",
+    "        idx_test  = split.test(i)\n",
     "\n",
     "        for j, k in enumerate(all_ks):\n",
     "            #print \"Round %d for k=%d...\" % (i, k)\n",

diff --git a/examples/meta/src/evaluation/cross_validation_timeseries_split.sg b/examples/meta/src/evaluation/cross_validation_timeseries_split.sg
@@ -11,6 +11,6 @@ TimeSeriesSplitting splitting(labels, num_subsets, min_subset_size)
 #![set parameters]
 
 #![generate subsets and inverse (aka test labels and train labels)]
-IntVector test_labels_indices = splitting.validation(1)
+IntVector test_labels_indices = splitting.test(1)
 IntVector train_labels_indices = splitting.train(1)
 #![generate subsets and inverse (aka test labels and train labels)]
diff --git a/examples/undocumented/libshogun/splitting_LOO_crossvalidation.cpp b/examples/undocumented/libshogun/splitting_LOO_crossvalidation.cpp
@@ -42,7 +42,7 @@ int main(int argc, char **argv)
 		{
 			//SG_SPRINT("subset %d\n", i);
 
-			SGVector<index_t> subset = splitting->validation(i);
+			SGVector<index_t> subset = splitting->test(i);
 			SGVector<index_t> inverse = splitting->train(i);
 
 			SGVector<index_t>::display_vector(subset.vector, subset.vlen, "subset indices");

diff --git a/examples/undocumented/libshogun/splitting_standard_crossvalidation.cpp b/examples/undocumented/libshogun/splitting_standard_crossvalidation.cpp
@@ -53,7 +53,7 @@ int main(int argc, char **argv)
 		{
 			SG_SPRINT("subset %d\n", i);
 
-			SGVector<index_t> subset = splitting->validation(i);
+			SGVector<index_t> subset = splitting->test(i);
 			SGVector<index_t> inverse = splitting->train(i);
 
 			SGVector<index_t>::display_vector(subset.vector, subset.vlen, "subset indices");

diff --git a/examples/undocumented/libshogun/splitting_stratified_crossvalidation.cpp b/examples/undocumented/libshogun/splitting_stratified_crossvalidation.cpp
@@ -54,7 +54,7 @@ int main(int argc, char **argv)
 
 		for (index_t i=0; i<num_subsets; ++i)
 		{
-			SGVector<index_t> subset = splitting->validation(i);
+			SGVector<index_t> subset = splitting->test(i);
 			SGVector<index_t> inverse = splitting->train(i);
 
 			SG_SPRINT("subset %d\n", i);
@@ -76,7 +76,7 @@ int main(int argc, char **argv)
 			SG_SPRINT("checking class %d\n", i);
 
 			/* count number of elements for this class */
-			SGVector<index_t> temp = splitting->validation(0);
+			SGVector<index_t> temp = splitting->test(0);
 			int32_t count=0;
 			for (index_t j=0; j<temp.vlen; ++j)
 			{
@@ -87,7 +87,7 @@ int main(int argc, char **argv)
 			/* check all subsets for same ratio */
 			for (index_t j=0; j<num_subsets; ++j)
 			{
-				SGVector<index_t> subset = splitting->validation(j);
+				SGVector<index_t> subset = splitting->test(j);
 				int32_t temp_count=0;
 				for (index_t k=0; k<subset.vlen; ++k)
 				{

diff --git a/src/shogun/evaluation/CrossValidation.cpp b/src/shogun/evaluation/CrossValidation.cpp
@@ -177,8 +177,7 @@ float64_t CCrossValidation::evaluate_one_run(
 			m_machine->train_locked(inverse_subset_indices);
 
 			/* feature subset for testing */
-			SGVector<index_t> subset_indices =
-			    m_splitting_strategy->validation(i);
+			SGVector<index_t> subset_indices = m_splitting_strategy->test(i);
 
 			/* evtl. update xvalidation output class */
 			fold->set_train_indices(inverse_subset_indices);
@@ -298,8 +297,7 @@ float64_t CCrossValidation::evaluate_one_run(
 
 			/* set feature subset for testing (subset method that stores
 			 * pointer) */
-			SGVector<index_t> subset_indices =
-			    m_splitting_strategy->validation(i);
+			SGVector<index_t> subset_indices = m_splitting_strategy->test(i);
 			features->add_subset(subset_indices);
 
 			/* set label subset for testing */

diff --git a/src/shogun/evaluation/CrossValidationSplitting.h b/src/shogun/evaluation/CrossValidationSplitting.h
@@ -43,7 +43,6 @@ class CCrossValidationSplitting: public CSplittingStrategy
 	CRandom * m_rng;
 
 protected:
-	/** implementation of the standard cross-validation splitting strategy */
 	virtual void build_subsets();
 };
 }

diff --git a/src/shogun/evaluation/SplittingStrategy.cpp b/src/shogun/evaluation/SplittingStrategy.cpp
@@ -69,7 +69,7 @@ CSplittingStrategy::~CSplittingStrategy()
 	SG_UNREF(m_subset_indices);
 }
 
-const SGVector<index_t> CSplittingStrategy::validation(index_t subset_idx) const
+const SGVector<index_t> CSplittingStrategy::test(index_t subset_idx) const
 {
 	if (!m_is_filled)
 	{

diff --git a/src/shogun/evaluation/SplittingStrategy.h b/src/shogun/evaluation/SplittingStrategy.h
@@ -21,7 +21,7 @@ class CLabels;
 
 /** @brief Abstract base class for all splitting types.
  * Takes a CLabels instance and generates a desired number of subsets which
- * can be accessed by their indices via the methods train() and validation().
+ * can be accessed by their indices via the methods train() and test().
  *
  * When being extended, the abstract method build_subsets() has to be
  * implemented. Its implementations must call reset_subsets() before and set
@@ -59,7 +59,7 @@ class CSplittingStrategy: public CSGObject
 	 * @return newly created vector of subset indices of the specified
 	 * subset is written here.
 	 */
-	const SGVector<index_t> validation(index_t subset_idx) const;
+	const SGVector<index_t> test(index_t subset_idx) const;
 
 	/** generates a newly created SGVector<index_t> with inverse indices of the
 	 * subset with the desired index. inverse here means all other indices.

diff --git a/src/shogun/multiclass/tree/RelaxedTreeUtil.cpp b/src/shogun/multiclass/tree/RelaxedTreeUtil.cpp
@@ -33,7 +33,7 @@ SGMatrix<float64_t> RelaxedTreeUtil::estimate_confusion_matrix(CBaseMulticlassMa
 		Y->remove_subset();
 
 		// subset for predicting
-		SGVector<index_t> subset_indices = split->validation(i);
+		SGVector<index_t> subset_indices = split->test(i);
 		X->add_subset(subset_indices);
 		Y->add_subset(subset_indices);
 

diff --git a/tests/unit/evaluation/SplittingStrategy_unittest.cc b/tests/unit/evaluation/SplittingStrategy_unittest.cc
@@ -46,7 +46,7 @@ TEST(SplittingStrategy,standard)
 
 		for (index_t i=0; i<num_subsets; ++i)
 		{
-			SGVector<index_t> subset = splitting->validation(i);
+			SGVector<index_t> subset = splitting->test(i);
 			SGVector<index_t> inverse = splitting->train(i);
 
 			for(index_t j=0;j<subset.vlen;++j)
@@ -126,7 +126,7 @@ TEST(SplittingStrategy,stratified_subsets_disjoint_cover)
 
 		for (index_t i=0; i<num_subsets; ++i)
 		{
-			SGVector<index_t> subset = splitting->validation(i);
+			SGVector<index_t> subset = splitting->test(i);
 			SGVector<index_t> inverse = splitting->train(i);
 
 			for(index_t j=0;j<subset.vlen;++j)
@@ -199,7 +199,7 @@ TEST(SplittingStrategy,stratified_subset_label_ratio)
 		for (index_t i=0; i<num_classes; ++i)
 		{
 			/* count number of elements for this class */
-			SGVector<index_t> temp = splitting->validation(0);
+			SGVector<index_t> temp = splitting->test(0);
 			int32_t count=0;
 			int32_t total_count=0;
 			for (index_t j=0; j<temp.vlen; ++j)
@@ -211,7 +211,7 @@ TEST(SplittingStrategy,stratified_subset_label_ratio)
 			/* check all subsets for same ratio */
 			for (index_t j=0; j<num_subsets; ++j)
 			{
-				SGVector<index_t> subset = splitting->validation(j);
+				SGVector<index_t> subset = splitting->test(j);
 				int32_t temp_count=0;
 				for (index_t k=0; k<subset.vlen; ++k)
 				{
@@ -255,7 +255,7 @@ TEST(SplittingStrategy,LOO)
 
 		for (index_t i=0; i<num_labels; ++i)
 		{
-			SGVector<index_t> subset = splitting->validation(i);
+			SGVector<index_t> subset = splitting->test(i);
 			SGVector<index_t> inverse = splitting->train(i);
 
 			for(index_t j=0;j<subset.vlen;++j)
@@ -310,7 +310,7 @@ TEST(SplittingStrategy, timeseries_subset_linear_splits)
 
 		for (index_t i = 0; i < num_subsets; ++i)
 		{
-			SGVector<index_t> subset = splitting->validation(i);
+			SGVector<index_t> subset = splitting->test(i);
 			SGVector<index_t> inverse = splitting->train(i);
 
 			/* Subset size should be atleat min_subset_size */
@@ -346,7 +346,7 @@ TEST(SplittingStrategy, timeseries_subsets_future_leak)
 
 		for (index_t i = 0; i < num_subsets; ++i)
 		{
-			SGVector<index_t> subset = splitting->validation(i);
+			SGVector<index_t> subset = splitting->test(i);
 			SGVector<index_t> inverse = splitting->train(i);
 
 			/* check future leak into test set */