Updated cross validation examples to use new API (#4309)

shogun-toolbox · Jun 1, 2018 · e1f02a5 · e1f02a5
1 parent 20f7808
commit e1f02a5
Show file tree

Hide file tree

Showing 5 changed files with 33 additions and 35 deletions.
diff --git a/examples/meta/src/evaluation/cross_validation_multiple_kernel_learning_weights_storage.sg b/examples/meta/src/evaluation/cross_validation_multiple_kernel_learning_weights_storage.sg
@@ -1,36 +1,34 @@
-CSVFile f_feats("../../data/fm_train_real.dat")
-CSVFile f_labels("../../data/label_train_twoclass.dat")
+File f_feats = csv_file("../../data/fm_train_real.dat")
+File f_labels = csv_file("../../data/label_train_twoclass.dat")
 
 #![create_features]
 Features feats = features(f_feats)
 CombinedFeatures combined_features()
 combined_features.append_feature_obj(feats)
 combined_features.append_feature_obj(feats)
 combined_features.append_feature_obj(feats)
-BinaryLabels labels(f_labels)
+Labels labels_train = labels(f_labels)
 #![create_features]
 
 #![create_kernels]
-CombinedKernel kernel()
-GaussianKernel k_1(10, 0.1)
-GaussianKernel k_2(10, 1)
-GaussianKernel k_3(10, 2)
-kernel.append_kernel(k_1)
-kernel.append_kernel(k_2)
-kernel.append_kernel(k_3)
+Kernel combined_kernel = kernel("CombinedKernel")
+Kernel k_1 = kernel("GaussianKernel", log_width=0.1)
+Kernel k_2 = kernel("GaussianKernel", log_width=1.0)
+Kernel k_3 = kernel("GaussianKernel", log_width=2.0)
+combined_kernel.add("kernel_array", k_1)
+combined_kernel.add("kernel_array", k_2)
+combined_kernel.add("kernel_array", k_3)
 #![create_kernels]
 
 #![create_classifier]
-LibSVM libsvm()
-MKLClassification svm(libsvm)
-svm.set_interleaved_optimization_enabled(False)
-svm.set_kernel(kernel)
+Machine libsvm = machine("LibSVM")
+Machine svm = machine("MKLClassification", kernel=combined_kernel, interleaved_optimization=False, svm=libsvm)
 #![create_classifier]
 
 #![create_cross_validation]
-StratifiedCrossValidationSplitting splitting_strategy(labels, 5)
-AccuracyMeasure evaluation_criterium()
-CrossValidation cross(svm, combined_features, labels, splitting_strategy, evaluation_criterium)
+StratifiedCrossValidationSplitting splitting_strategy(labels_train, 5)
+Evaluation evaluation_criterium = evaluation("AccuracyMeasure")
+CrossValidation cross(svm, combined_features, labels_train, splitting_strategy, evaluation_criterium)
 cross.set_autolock(False)
 cross.set_num_runs(2)
 #![create_cross_validation]
@@ -52,10 +50,11 @@ real stddev = result.get_real("std_dev")
 #![get_fold_machine]
 CrossValidationStorage obs = mkl_obs.get_observation(0)
 CrossValidationFoldStorage fold = obs.get_fold(0)
-MKLClassification machine = MKLClassification:obtain_from_generic(fold.get_trained_machine())
+Machine machine_mkl = machine(fold.get_trained_machine())
+
 #![get_fold_machine]
 
 #![get_weights]
-CombinedKernel k = CombinedKernel:obtain_from_generic(machine.get_kernel())
+Kernel k = kernel(machine_mkl.get("kernel"))
 RealVector w = k.get_subkernel_weights()
 #![get_weights]
diff --git a/examples/meta/src/evaluation/cross_validation_support_vector_machine.sg b/examples/meta/src/evaluation/cross_validation_support_vector_machine.sg
@@ -1,5 +1,5 @@
-CSVFile f_feats("../../data/fm_train_real.dat")
-CSVFile f_labels("../../data/label_train_twoclass.dat")
+File f_feats = csv_file("../../data/fm_train_real.dat")
+File f_labels = csv_file("../../data/label_train_twoclass.dat")
 
 Features feats = features(f_feats)
 Labels labs = labels(f_labels)
@@ -8,11 +8,11 @@ Machine svm = machine("LibLinear")
 svm.put("liblinear_solver_type", enum LIBLINEAR_SOLVER_TYPE.L2R_L2LOSS_SVC)
 
 StratifiedCrossValidationSplitting splitting_strategy(labs, 5)
-AccuracyMeasure evaluation_criterion()
+Evaluation evaluation_criterion = evaluation("AccuracyMeasure")
 CrossValidation cross(svm, feats, labs, splitting_strategy, evaluation_criterion)
 cross.set_num_runs(2)
 
 CrossValidationResult result = CrossValidationResult:obtain_from_generic(cross.evaluate())
 
-real mean = result.get_mean()
-real stddev = result.get_std_dev()
+real mean = result.get_real("mean")
+real stddev = result.get_real("std_dev")
diff --git a/examples/meta/src/evaluation/cross_validation_timeseries_split.sg b/examples/meta/src/evaluation/cross_validation_timeseries_split.sg
@@ -1,13 +1,13 @@
-CSVFile f_labels("../../data/label_train_regression.dat")
+File f_labels = csv_file("../../data/label_train_regression.dat")
 
 #![create labels]
-RegressionLabels labels(f_labels)
+Labels reg_labels = labels(f_labels)
 #![create labels]
 
 #![set parameters]
 int num_subsets = 5
 int min_subset_size = 4
-TimeSeriesSplitting splitting(labels, num_subsets)
+TimeSeriesSplitting splitting(reg_labels, num_subsets)
 splitting.set_min_subset_size(min_subset_size)
 #![set parameters]
 
@@ -18,4 +18,4 @@ splitting.build_subsets()
 #![generate subsets and inverse (aka test labels and train labels)]
 IntVector test_labels_indices = splitting.generate_subset_indices(1)
 IntVector train_labels_indices = splitting.generate_subset_inverse(1)
-#![generate subsets and inverse (aka test labels and train labels)]
+#![generate subsets and inverse (aka test labels and train labels)]
diff --git a/examples/meta/src/regression/least_angle_regression.sg b/examples/meta/src/regression/least_angle_regression.sg
@@ -40,4 +40,3 @@ real mse = eval.evaluate(labels_predict, labels_test)
 
 # integration testing variables
 RealVector output = labels_predict.get_real_vector("labels")
-
diff --git a/examples/meta/src/regression/random_forest_regression.sg b/examples/meta/src/regression/random_forest_regression.sg
@@ -1,7 +1,7 @@
-CSVFile f_feats_train("../../data/regression_1d_linear_features_train.dat")
-CSVFile f_feats_test("../../data/regression_1d_linear_features_test.dat")
-CSVFile f_labels_train("../../data/regression_1d_linear_labels_train.dat")
-CSVFile f_labels_test("../../data/regression_1d_linear_labels_test.dat")
+File f_feats_train    = csv_file("../../data/regression_1d_linear_features_train.dat")
+File f_feats_test     = csv_file("../../data/regression_1d_linear_features_test.dat")
+File f_labels_train   = csv_file("../../data/regression_1d_linear_labels_train.dat")
+File f_labels_test    = csv_file("../../data/regression_1d_linear_labels_test.dat")
 Math:init_random(1)
 
 #![create_features]
@@ -26,10 +26,10 @@ RegressionLabels labels_predict = rand_forest.apply_regression(features_test)
 #![train_and_apply]
 
 #![evaluate_error]
-MeanSquaredError mse()
+Evaluation mse = evaluation("MeanSquaredError")
 real oob = rand_forest.get_oob_error(mse)
 real mserror = mse.evaluate(labels_predict, labels_test)
 #![evaluate_error]
 
 # additional integration testing variables
-RealVector output = labels_predict.get_labels()
+RealVector output = labels_predict.get_real_vector("labels")