Migrating GAN related tests and math tests to Catch2

src/mlpack/tests/CMakeLists.txt

@@ -5,12 +5,10 @@ add_executable(mlpack_test
   callback_test.cpp
   cf_test.cpp
   cosine_tree_test.cpp
-  dcgan_test.cpp
   drusilla_select_test.cpp
   emst_test.cpp
   fastmks_test.cpp
   facilities_test.cpp
-  gan_test.cpp
   gmm_test.cpp
   hmm_test.cpp
   hoeffding_tree_test.cpp
@@ -26,7 +24,6 @@ add_executable(mlpack_test
   log_test.cpp
   logistic_regression_test.cpp
   lsh_test.cpp
-  math_test.cpp
   matrix_completion_test.cpp
   maximal_inputs_test.cpp
   mlpack_test.cpp
@@ -59,7 +56,6 @@ add_executable(mlpack_test
   ub_tree_test.cpp
   union_find_test.cpp
   vantage_point_tree_test.cpp
-  wgan_test.cpp
   main_tests/cf_test.cpp
   main_tests/det_test.cpp
   main_tests/emst_test.cpp
@@ -108,11 +104,13 @@ add_executable(mlpack_catch_test
   convolution_test.cpp
   cv_test.cpp
   dbscan_test.cpp
+  dcgan_test.cpp
   decision_stump_test.cpp
   decision_tree_test.cpp
   det_test.cpp
   distribution_test.cpp
   feedforward_network_test.cpp
+  gan_test.cpp
   image_load_test.cpp
   imputation_test.cpp
   io_test.cpp
@@ -129,6 +127,7 @@ add_executable(mlpack_catch_test
   load_save_test.cpp
   loss_functions_test.cpp
   main.cpp
+  math_test.cpp
   metric_test.cpp
   mean_shift_test.cpp
   nca_test.cpp
@@ -155,6 +154,7 @@ add_executable(mlpack_catch_test
   svd_incremental_test.cpp
   svdplusplus_test.cpp
   test_catch_tools.hpp
+  wgan_test.cpp
   main_tests/adaboost_test.cpp
   main_tests/approx_kfn_test.cpp
   main_tests/bayesian_linear_regression_test.cpp

src/mlpack/tests/dcgan_test.cpp

@@ -20,8 +20,7 @@
 
 #include <ensmallen.hpp>
 
-#include <boost/test/unit_test.hpp>
-#include "test_tools.hpp"
+#include "catch.hpp"
 #include "serialization.hpp"
 
 using namespace mlpack;
@@ -30,14 +29,12 @@ using namespace mlpack::math;
 using namespace mlpack::regression;
 using namespace std::placeholders;
 
-BOOST_AUTO_TEST_SUITE(DCGANNetworkTest);
-
 /*
  * Tests the DCGAN implementation on the MNIST dataset.
  * It's not viable to train on bigger parameters due to time constraints.
  * Please refer mlpack/models repository for the tutorial.
  */
-BOOST_AUTO_TEST_CASE(DCGANMNISTTest)
+TEST_CASE("DCGANMNISTTest", "[DCGANNetworkTest]")
 {
   size_t dNumKernels = 32;
   size_t discriminatorPreTrain = 5;
@@ -129,7 +126,7 @@ BOOST_AUTO_TEST_CASE(DCGANMNISTTest)
   double objVal = dcgan.Train(trainData, optimizer);
 
   // Test that objective value returned by GAN::Train() is finite.
-  BOOST_REQUIRE_EQUAL(std::isfinite(objVal), true);
+  REQUIRE(std::isfinite(objVal) == true);
 
   // Generate samples.
   Log::Info << "Sampling..." << std::endl;
@@ -196,7 +193,7 @@ BOOST_AUTO_TEST_CASE(DCGANMNISTTest)
  * Tests the DCGAN implementation with minibatch layer on the MNIST dataset.
  * It's not viable to train on bigger parameters due to time constraints.
 
-BOOST_AUTO_TEST_CASE(DCGANMNISTTest)
+TEST_CASE("DCGANMNISTTest", "[DCGANNetworkTest]")
 {
   size_t dNumKernels = 32;
   size_t discriminatorPreTrain = 5;
@@ -284,7 +281,7 @@ BOOST_AUTO_TEST_CASE(DCGANMNISTTest)
   double objVal = dcgan.Train(optimizer);
 
   // Test that objective value returned by GAN::Train() is finite.
-  BOOST_REQUIRE_EQUAL(std::isfinite(objVal), true);
+  REQUIRE(std::isfinite(objVal) == true);
 
   // Generate samples
   Log::Info << "Sampling..." << std::endl;
@@ -350,7 +347,7 @@ BOOST_AUTO_TEST_CASE(DCGANMNISTTest)
  * It's currently not possible to run this every time due to time constraints.
  * Please refer mlpack/models repository for the tutorial.
 
-BOOST_AUTO_TEST_CASE(DCGANCelebATest)
+TEST_CASE("DCGANCelebATest", "[DCGANNetworkTest]")
 {
   size_t dNumKernels = 64;
   size_t discriminatorPreTrain = 300;
@@ -470,5 +467,3 @@ BOOST_AUTO_TEST_CASE(DCGANCelebATest)
   Log::Info << "Output generated!" << std::endl;
 }
 */
-
-BOOST_AUTO_TEST_SUITE_END();

src/mlpack/tests/gan_test.cpp

@@ -20,8 +20,7 @@
 
 #include <ensmallen.hpp>
 
-#include <boost/test/unit_test.hpp>
-#include "test_tools.hpp"
+#include "catch.hpp"
 #include "serialization.hpp"
 
 using namespace mlpack;
@@ -30,14 +29,12 @@ using namespace mlpack::math;
 using namespace mlpack::regression;
 using namespace std::placeholders;
 
-BOOST_AUTO_TEST_SUITE(GANNetworkTest);
-
 /*
  * Load pre trained network values
  * for generating distribution that
  * is close to N(4, 0.5)
  */
-BOOST_AUTO_TEST_CASE(GANTest)
+TEST_CASE("GANTest", "[GANNetworkTest]")
 {
   size_t generatorHiddenLayerSize = 8;
   size_t discriminatorHiddenLayerSize = 8;
@@ -125,8 +122,8 @@ BOOST_AUTO_TEST_CASE(GANTest)
   double originalStd = arma::as_scalar(arma::stddev(
       generatedData.rows(dim, 2 * dim - 1), 0, 1));
 
-  BOOST_REQUIRE_LE(generatedMean - originalMean, 0.2);
-  BOOST_REQUIRE_LE(generatedStd - originalStd, 0.2);
+  REQUIRE(generatedMean - originalMean <= 0.2);
+  REQUIRE(generatedStd - originalStd <= 0.2);
 }
 
 /*
@@ -134,7 +131,7 @@ BOOST_AUTO_TEST_CASE(GANTest)
  * It's not viable to train on bigger parameters due to time constraints.
  * Please refer mlpack/models repository for the tutorial.
  */
-BOOST_AUTO_TEST_CASE(GANMNISTTest)
+TEST_CASE("GANMNISTTest", "[GANNetworkTest]")
 {
   size_t dNumKernels = 32;
   size_t discriminatorPreTrain = 5;
@@ -217,8 +214,8 @@ BOOST_AUTO_TEST_CASE(GANMNISTTest)
   Log::Info << "Training..." << std::endl;
   std::stringstream stream;
   double objVal = gan.Train(trainData, optimizer, ens::ProgressBar(70, stream));
-  BOOST_REQUIRE_GT(stream.str().length(), 0);
-  BOOST_REQUIRE_EQUAL(std::isfinite(objVal), true);
+  REQUIRE(stream.str().length() > 0);
+  REQUIRE(std::isfinite(objVal) == true);
 
   // Generate samples.
   Log::Info << "Sampling..." << std::endl;
@@ -284,7 +281,7 @@ BOOST_AUTO_TEST_CASE(GANMNISTTest)
  * Create GAN network and test for memory sharing
  * between discriminator and gan predictors.
  */
-BOOST_AUTO_TEST_CASE(GANMemorySharingTest)
+TEST_CASE("GANMemorySharingTest", "[GANNetworkTest]")
 {
   size_t generatorHiddenLayerSize = 8;
   size_t discriminatorHiddenLayerSize = 8;
@@ -347,5 +344,3 @@ BOOST_AUTO_TEST_CASE(GANMemorySharingTest)
   CheckMatricesNotEqual(gan.Predictors().head_cols(trainData.n_cols),
       trainData);
 }
-
-BOOST_AUTO_TEST_SUITE_END();

src/mlpack/tests/wgan_test.cpp

@@ -20,8 +20,7 @@
 
 #include <ensmallen.hpp>
 
-#include <boost/test/unit_test.hpp>
-#include "test_tools.hpp"
+#include "catch.hpp"
 #include "serialization.hpp"
 
 using namespace mlpack;
@@ -30,14 +29,12 @@ using namespace mlpack::math;
 using namespace mlpack::regression;
 using namespace std::placeholders;
 
-BOOST_AUTO_TEST_SUITE(WGANNetworkTest);
-
 /*
  * Tests the standard WGAN implementation on the MNIST dataset.
  * It's not viable to train on bigger parameters due to time constraints.
  * Please refer mlpack/models repository for the tutorial.
  */
-BOOST_AUTO_TEST_CASE(WGANMNISTTest)
+TEST_CASE("WGANMNISTTest", "[WGANNetworkTest]")
 {
   size_t dNumKernels = 32;
   size_t discriminatorPreTrain = 5;
@@ -131,7 +128,7 @@ BOOST_AUTO_TEST_CASE(WGANMNISTTest)
   double objVal = wgan.Train(trainData, optimizer);
 
   // Test that objective value returned by GAN::Train() is finite.
-  BOOST_REQUIRE_EQUAL(std::isfinite(objVal), true);
+  REQUIRE(std::isfinite(objVal) == true);
 
   // Generate samples.
   Log::Info << "Sampling..." << std::endl;
@@ -198,7 +195,7 @@ BOOST_AUTO_TEST_CASE(WGANMNISTTest)
  * It's not viable to train on bigger parameters due to time constraints.
  * Please refer mlpack/models repository for the tutorial.
  */
-BOOST_AUTO_TEST_CASE(WGANGPMNISTTest)
+TEST_CASE("WGANGPMNISTTest","[WGANNetworkTest]")
 {
   size_t dNumKernels = 32;
   size_t discriminatorPreTrain = 5;
@@ -293,7 +290,7 @@ BOOST_AUTO_TEST_CASE(WGANGPMNISTTest)
   double objVal = wganGP.Train(trainData, optimizer);
 
   // Test that objective value returned by GAN::Train() is finite.
-  BOOST_REQUIRE_EQUAL(std::isfinite(objVal), true);
+  REQUIRE(std::isfinite(objVal) == true);
 
   // Generate samples.
   Log::Info << "Sampling..." << std::endl;
@@ -354,5 +351,3 @@ BOOST_AUTO_TEST_CASE(WGANGPMNISTTest)
   CheckMatrices(orgPredictions, textPredictions);
   CheckMatrices(orgPredictions, binaryPredictions);
 }
-
-BOOST_AUTO_TEST_SUITE_END();