clean

Artelnics · Dec 10, 2024 · 3e7534b · 3e7534b
1 parent ede684c
commit 3e7534b
Show file tree

Hide file tree

Showing 5 changed files with 37 additions and 169 deletions.
diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
@@ -7,24 +7,35 @@ on:
     types: [opened, synchronize, reopened]
     branches:
       - dev
+
 jobs:
   build:
     name: Build
     runs-on: windows-latest
     env:
-      BUILD_WRAPPER_OUT_DIR: build_wrapper_output_directory # Directory where build-wrapper output will be placed
+      BUILD_WRAPPER_OUT_DIR: build_wrapper_output_directory
     steps:
       - uses: actions/checkout@v4
         with:
-          fetch-depth: 0 # Shallow clones should be disabled for a better relevancy of analysis
+          fetch-depth: 0
       - name: Set up msbuild
         uses: microsoft/setup-msbuild@v2
       - name: Install sonar-scanner and build-wrapper
         uses: sonarsource/sonarcloud-github-c-cpp@v3
+
+      - name: Build with build-wrapper
+        run: |
+          build-wrapper-win-x86-64.exe --out-dir build_wrapper_output_directory msbuild opennn.sln /t:Rebuild /p:Configuration=Release
+
       - name: Run sonar-scanner
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          SONAR_TOKEN: ${{ secrets.SONARCLOUD_TOKEN }} # Put the name of your token here
+          SONAR_TOKEN: ${{ secrets.SONARCLOUD_TOKEN }}
         run: |
           sonar-scanner \
-            --define sonar.inclusions="opennn/opennn/**/*.cpp,opennn/opennn/**/*.h"
+            -Dsonar.projectKey=opennn \
+            -Dsonar.organization=Artelnics \
+            -Dsonar.host.url=https://sonarcloud.io \
+            -Dsonar.login=${{ secrets.SONARCLOUD_TOKEN }} \
+            -Dsonar.cfamily.build-wrapper-output=build_wrapper_output_directory \
+            -Dsonar.inclusions="opennn/opennn/**/*.cpp,opennn/opennn/**/*.h"
diff --git a/examples/mnist/main.cpp b/examples/mnist/main.cpp
@@ -53,7 +53,7 @@ int main()
 
         NeuralNetwork neural_network(NeuralNetwork::ModelType::ImageClassification,
             image_data_set.get_input_dimensions(),
-            { 16,8 },
+            { 16 },
             image_data_set.get_target_dimensions());
 
         //neural_network.print();

diff --git a/examples/mnist/mnist.vcxproj b/examples/mnist/mnist.vcxproj
@@ -124,6 +124,7 @@
       <ImportLibrary>Debug/mnist.lib</ImportLibrary>
       <ProgramDataBaseFile>Debug/mnist.pdb</ProgramDataBaseFile>
       <SubSystem>Console</SubSystem>
+      <Profile>true</Profile>
     </Link>
     <ProjectReference>
       <LinkLibraryDependencies>false</LinkLibraryDependencies>
@@ -176,6 +177,7 @@
       <ImportLibrary>Release/mnist.lib</ImportLibrary>
       <ProgramDataBaseFile>Release/mnist.pdb</ProgramDataBaseFile>
       <SubSystem>Console</SubSystem>
+      <Profile>true</Profile>
     </Link>
     <ProjectReference>
       <LinkLibraryDependencies>false</LinkLibraryDependencies>
@@ -221,6 +223,7 @@
       <ImportLibrary>MinSizeRel/mnist.lib</ImportLibrary>
       <ProgramDataBaseFile>MinSizeRel/mnist.pdb</ProgramDataBaseFile>
       <SubSystem>Console</SubSystem>
+      <Profile>true</Profile>
     </Link>
     <ProjectReference>
       <LinkLibraryDependencies>false</LinkLibraryDependencies>
@@ -265,6 +268,7 @@
       <ImportLibrary>RelWithDebInfo/mnist.lib</ImportLibrary>
       <ProgramDataBaseFile>RelWithDebInfo/mnist.pdb</ProgramDataBaseFile>
       <SubSystem>Console</SubSystem>
+      <Profile>true</Profile>
     </Link>
     <ProjectReference>
       <LinkLibraryDependencies>false</LinkLibraryDependencies>

diff --git a/tests/convolutional_layer_test.cpp b/tests/convolutional_layer_test.cpp
@@ -1,149 +1,21 @@
 #include "pch.h"
 
 #include "../opennn/convolutional_layer.h"
-/*
-TEST(ConvolutionalLayerTest, EigenConvolution)
-{
-    Tensor<type, 2> input_2;
-    Tensor<type, 2> kernel_2;
-    Tensor<type, 2> output_2;
-
-    std::cout << "Hello";
-
-    const Eigen::array<ptrdiff_t, 2> dimensions_2 = { 0, 1 };
-
-    Tensor<type, 3> input_3;
-    Tensor<type, 3> kernel_3;
-    Tensor<type, 3> output_3;
-
-    const Eigen::array<ptrdiff_t, 3> dimensions_3 = { 0, 1, 2 };
-
-    Tensor<type, 4> input_4;
-    Tensor<type, 3> kernel_4;
-    Tensor<type, 4> output_4;
-
-    const Eigen::array<ptrdiff_t, 3> dimensions_4 = { 1, 2, 3 };
-
-    // Convolution 2D, 1 channel
-
-    input_2.resize(3, 3);
-    input_2.setRandom();
-
-    kernel_2.resize(2, 2);
-    kernel_2.setRandom();
-
-    output_2 = input_2.convolve(kernel_2, dimensions_2);
-
-//    EXPECT_EQ(output_2.dimension(0), dimensions{ 2,1 });
-
-//    EXPECT_EQ( == 2);
-//    EXPECT_EQ(output_2.dimension(1) == 2);
-
-    // Convolution 3D, 3 channels
-
-    input_3.resize(5, 5, 3);
-    input_3.setRandom();
-
-    kernel_3.resize(2, 2, 3);
-    kernel_3.setRandom();
-
-    output_3 = input_3.convolve(kernel_3, dimensions_3);
-
-//    EXPECT_EQ(output_3.dimension(0) == 4);
-//    EXPECT_EQ(output_3.dimension(1) == 4);
-//    EXPECT_EQ(output_3.dimension(2) == 1);
-
-    // Convolution 2D, 3 channels, multiple images, 1 kernel
-
-    input_4.resize(10, 3, 5, 5);
-    input_4.setConstant(type(1));
-    input_4.chip(1, 0).setConstant(type(2));
-    input_4.chip(2, 0).setConstant(type(3));
-
-    kernel_4.resize(3, 2, 2);
-
-    kernel_4.setConstant(type(1.0 / 12.0));
-
-    output_4 = input_4.convolve(kernel_4, dimensions_4);
-
-//    EXPECT_EQ(output_3.dimension(0) == 10);
-//    EXPECT_EQ(output_3.dimension(1) == 1);
-//    EXPECT_EQ(output_3.dimension(2) == 4);
-//    EXPECT_EQ(output_3.dimension(3) == 4);
-
-    // EXPECT_EQ(abs(output_3(0, 0, 0, 0) - type(1)) < type(NUMERIC_LIMITS_MIN)
-    //                 && abs(output_3(0, 0, 0, 1) - type(1)) < type(NUMERIC_LIMITS_MIN)
-    //                 && abs(output_3(0, 0, 0, 2) - type(1)) < type(NUMERIC_LIMITS_MIN)
-    //                 && abs(output_3(0, 0, 0, 3) - type(1)) < type(NUMERIC_LIMITS_MIN)
-    //                 && abs(output_3(0, 0, 1, 0) - type(1)) < type(NUMERIC_LIMITS_MIN)
-    //                 && abs(output_3(0, 0, 1, 1) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 1, 2) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 1, 3) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 2, 0) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 2, 1) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 2, 2) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 2, 3) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 3, 0) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 3, 1) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 3, 2) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(0, 0, 3, 3) - type(1)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 0, 0) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 0, 1) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 0, 2) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 0, 3) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 1, 0) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 1, 1) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 1, 2) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 1, 3) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 2, 0) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 2, 1) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 2, 2) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 2, 3) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 3, 0) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 3, 1) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 3, 2) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(1, 0, 3, 3) - type(2)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 0, 0) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 0, 1) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 0, 2) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 0, 3) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 1, 0) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 1, 1) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 1, 2) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 1, 3) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 2, 0) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 2, 1) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 2, 2) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 2, 3) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 3, 0) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 3, 1) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 3, 2) - type(3)) < type(NUMERIC_LIMITS_MIN) &&
-    //             abs(output_3(2, 0, 3, 3) - type(3)) <= type(NUMERIC_LIMITS_MIN));
-
-    // Convolution 3D, 2 channels
-
-    Tensor<type, 3> input(3, 3, 2);
-    Tensor<type, 3> kernel(2, 2, 2);
-    Tensor<type, 3> output(2, 2, 1);
-
-    for (int i = 0; i < 3 * 3 * 2; i++)
-        *(input.data() + i) = i;
-
-    for (int i = 0; i < 2 * 2 * 2; i++)
-        *(kernel.data() + i) = i + 1;
-
-    const Eigen::array<ptrdiff_t, 3> dimensions = { 0, 1, 2 };
-
-    output = input.convolve(kernel, dimensions);
-
-//    EXPECT_EQ(fabs(output(0, 0, 0) - 320) < type(NUMERIC_LIMITS_MIN));
-//    EXPECT_EQ(fabs(output(1, 0, 0) - 356) < type(NUMERIC_LIMITS_MIN));
-//    EXPECT_EQ(fabs(output(0, 1, 0) - 428) < type(NUMERIC_LIMITS_MIN));
-//    EXPECT_EQ(fabs(output(1, 1, 0) - 464) < type(NUMERIC_LIMITS_MIN));
-}
-=======
 #include "../opennn/tensors.h"
->>>>>>> acde2b25554c3cb1fe0ef447c58dfba1588857fa
+
+
+Tensor<type, 4> generate_input_tensor_convolution(const Tensor<type, 2>& data,
+    const vector<Index>& rows_indices,
+    const vector<Index>& columns_indices,
+    const dimensions& input_dimensions) {
+    Tensor<type, 4> input_tensor(rows_indices.size(),
+        input_dimensions[0],
+        input_dimensions[1],
+        input_dimensions[2]);
+    type* tensor_data = input_tensor.data();
+    fill_tensor_data(data, rows_indices, columns_indices, tensor_data);
+    return input_tensor;
+}
 
 
 struct ConvolutionalLayerConfig {
@@ -157,24 +29,8 @@ struct ConvolutionalLayerConfig {
     Tensor<type, 4> expected_output;
 };
 
-
 class ConvolutionalLayerTest : public ::testing::TestWithParam<ConvolutionalLayerConfig> {};
 
-
-Tensor<type, 4> generate_input_tensor_convolution(const Tensor<type, 2>& data,
-                                                  const vector<Index>& rows_indices,
-                                                  const vector<Index>& columns_indices,
-                                                  const dimensions& input_dimensions) {
-    Tensor<type, 4> input_tensor(rows_indices.size(),
-                                 input_dimensions[0],
-                                 input_dimensions[1],
-                                 input_dimensions[2]);
-    type* tensor_data = input_tensor.data();
-    fill_tensor_data(data, rows_indices, columns_indices, tensor_data);
-    return input_tensor;
-}
-
-
 INSTANTIATE_TEST_CASE_P(ConvolutionalLayerTests, ConvolutionalLayerTest, ::testing::Values(
     ConvolutionalLayerConfig{
         {4, 4, 1}, {3, 3, 1, 1}, {1, 1}, ConvolutionalLayer::ActivationFunction::Linear, ConvolutionalLayer::ConvolutionType::Valid, "ConvolutionLayer",
@@ -362,5 +218,4 @@ TEST_P(ConvolutionalLayerTest, BackPropagate) {
     EXPECT_EQ(weight_derivatives.dimension(1), parameters.kernel_dimensions[0]);
     EXPECT_EQ(weight_derivatives.dimension(2), parameters.kernel_dimensions[1]);
     EXPECT_EQ(weight_derivatives.dimension(3), parameters.kernel_dimensions[2]);
-}
-*/
+}
diff --git a/tests/test.cpp b/tests/test.cpp
@@ -9,11 +9,9 @@ int main(int argc, char **argv) {
 
     ::testing::InitGoogleTest(&argc, argv);
 
-    //::testing::GTEST_FLAG(filter) = "PerformanceTest.ImageClassification";
-
     //::testing::GTEST_FLAG(filter) = "FlattenLayerTest";
-    ::testing::GTEST_FLAG(filter) = "PoolingLayerTests/*";
-    //::testing::GTEST_FLAG(filter) = "ConvolutionalLayerTest/*";
+    //::testing::GTEST_FLAG(filter) = "PoolingLayerTests/*";
+    //::testing::GTEST_FLAG(filter) = "ConvolutionalLayerTests/*";
 
     return RUN_ALL_TESTS();
 }