working closer

CMakeLists.txt

@@ -1,5 +1,5 @@
 cmake_minimum_required(VERSION 3.25)
-project(COSC-4P80-Assignment-3 VERSION 0.0.17)
+project(COSC-4P80-Assignment-3 VERSION 0.0.18)
 include(FetchContent)
 
 option(ENABLE_ADDRSAN "Enable the address sanitizer" OFF)

include/assign3/functions.h

@@ -65,6 +65,18 @@ namespace assign3
         [[nodiscard]] Scalar distance(blt::span<const Scalar> x, blt::span<const Scalar> y) const final;
     };
     
+    struct toroidal_euclidean_distance_function_t : public distance_function_t
+    {
+        public:
+            toroidal_euclidean_distance_function_t(blt::i32 width, blt::i32 height): width(width), height(height)
+            {}
+            
+            [[nodiscard]] Scalar distance(blt::span<const Scalar> x, blt::span<const Scalar> y) const final;
+        
+        private:
+            blt::i32 width, height;
+    };
+    
 }
 
 #endif //COSC_4P80_ASSIGNMENT_3_FUNCTIONS_H

include/assign3/manager.h

@@ -84,6 +84,8 @@ namespace assign3
             
             std::vector<float> get_neuron_activations(const data_file_t& file);
             
+            static std::vector<float> normalize_data(const std::vector<float>& data);
+            
             void draw_som(neuron_render_info_t info, const std::function<blt::vec4(render_data_t)>& color_func);
             
             void draw_debug(const data_file_t& file);
@@ -116,9 +118,14 @@ namespace assign3
             
             int currently_selected_network = 0;
             bool debug_mode = false;
+            bool draw_colors = true;
+            bool draw_data_lines = false;
             bool running = false;
             int debug_state = 0;
             int selected_data_point = 0;
+            
+            float requested_activation = 0.5;
+            float at_distance_measurement = 2;
     };
     
 }

include/assign3/neuron.h

@@ -38,7 +38,7 @@ namespace assign3
             
             neuron_t& update(const std::vector<Scalar>& new_data, Scalar dist, Scalar eta);
             
-            static Scalar distance(const neuron_t& n1, const neuron_t& n2);
+            static Scalar distance(distance_function_t* dist_func, const neuron_t& n1, const neuron_t& n2);
             
             [[nodiscard]] Scalar dist(const std::vector<Scalar>& X) const;
             

src/functions.cpp

@@ -18,6 +18,7 @@
 #include <assign3/functions.h>
 #include <cmath>
 #include "blt/iterator/zip.h"
+#include <blt/std/assert.h>
 
 namespace assign3
 {
@@ -43,4 +44,14 @@ namespace assign3
         }
         return std::sqrt(dist);
     }
+    
+    Scalar toroidal_euclidean_distance_function_t::distance(blt::span<const Scalar> x, blt::span<const Scalar> y) const
+    {
+        BLT_ASSERT(x.size() == 2 && y.size() == 2);
+        Scalar x_diff = x[0] - y[0];
+        Scalar y_diff = x[1] - y[1];
+        Scalar x_min = std::min(x_diff, static_cast<Scalar>(width) - x_diff);
+        Scalar y_min = std::min(y_diff, static_cast<Scalar>(height) - y_diff);
+        return std::sqrt(x_min * x_min + y_min * y_min);
+    }
 }

src/main.cpp

@@ -58,6 +58,8 @@ int main(int argc, const char** argv)
     auto args = parser.parse_args(argc, argv);
     
     data.files = assign3::data_file_t::load_data_files_from_path(args.get<std::string>("file"));
+    for (auto& v : data.files)
+        v = v.normalize();
     data.update();
     
     blt::gfx::init(blt::gfx::window_data{"My Sexy Window", init, update, destroy}.setSyncInterval(1).setMaximized(true));

src/manager.cpp

@@ -20,6 +20,7 @@
 #include <blt/math/log_util.h>
 #include <imgui.h>
 #include <implot.h>
+#include <algorithm>
 
 static void HelpMarker(const std::string& desc)
 {
@@ -38,6 +39,23 @@ const char* get_selection_string(void* user_data, int selection)
     return (reinterpret_cast<std::string*>(user_data))[selection].c_str();
 }
 
+std::vector<assign3::Scalar> rotate90Clockwise(const std::vector<assign3::Scalar>& input, int width, int height)
+{
+    std::vector<assign3::Scalar> rotated(width * height);
+    
+    for (int row = 0; row < height; ++row)
+    {
+        for (int col = 0; col < width; ++col)
+        {
+            int newRow = col;
+            int newCol = height - row - 1;
+            rotated[newRow * height + newCol] = input[row * width + col];
+        }
+    }
+    
+    return rotated;
+}
+
 namespace assign3
 {
     
@@ -142,7 +160,9 @@ namespace assign3
                     {
                         case debug_type::DATA_POINT:
                         {
-                            auto current_data_file = motor_data.files[currently_selected_network].normalize();
+                            ImGui::Checkbox("Data Type Color", &draw_colors);
+                            ImGui::Checkbox("Data Lines", &draw_data_lines);
+                            auto current_data_file = motor_data.files[currently_selected_network];
                             std::vector<std::string> names;
                             for (const auto& [i, v] : blt::enumerate(current_data_file.data_points))
                                 names.push_back("#" + std::to_string(i) + " (" + (v.is_bad ? "Bad)" : "Good)"));
@@ -160,17 +180,34 @@ namespace assign3
         }
         ImGui::End();
         
+        auto current_data_file = motor_data.files[currently_selected_network];
+        auto closest_type = get_neuron_activations(current_data_file);
+        
+        if (ImGui::Begin("Plots & Data"))
+        {
+            ImPlot::SetNextAxesLimits(0, som_width, 0, som_height, ImPlotCond_Always);
+            if (ImPlot::BeginPlot("Activations", ImVec2(-1, 0), ImPlotFlags_NoInputs))
+            {
+                auto rev = rotate90Clockwise(closest_type, som_width, som_height);
+//                auto rev = closest_type;
+//                std::reverse(rev.begin(), rev.end());
+                ImPlot::PlotHeatmap("##data_map", rev.data(), som_height, som_width, 0, 0, "%.1f", ImPlotPoint(0, 0),
+                                    ImPlotPoint(som_width, som_height), ImPlotHeatmapFlags_ColMajor);
+                ImPlot::EndPlot();
+            }
+        }
+        ImGui::End();
+        
         if (running)
         {
             if (som->get_current_epoch() < som->get_max_epochs())
                 som->train_epoch(initial_learn_rate, topology_function.get());
         }
         
-        auto current_data_file = motor_data.files[currently_selected_network].normalize();
         
         if (!debug_mode)
         {
-            auto closest_type = get_neuron_activations(current_data_file);
+            closest_type = normalize_data(closest_type);
             draw_som(neuron_render_info_t{}.set_base_pos({370, 145}).set_neuron_scale(120).set_neuron_padding({5, 5}),
                      [&closest_type](render_data_t context) {
                          auto type = closest_type[context.index];
@@ -207,27 +244,20 @@ namespace assign3
         closest_type.clear();
         closest_type.resize(som->get_array().get_map().size());
         
+        
         for (auto [i, v] : blt::enumerate(som->get_array().get_map()))
         {
-            auto half = som->find_closest_neighbour_distance(i);
+            auto half = som->find_closest_neighbour_distance(i) / at_distance_measurement;
-            auto scale = topology_function->scale(half * 0.5f, 0.5);
+            auto scale = topology_function->scale(half, requested_activation);
             for (const auto& data : file.data_points)
             {
-                auto dist = v.dist(data.bins);
+                auto ds = topology_function->call(v.dist(data.bins), scale);
-                auto ds = topology_function->call(dist, scale);
                 if (data.is_bad)
                     closest_type[i] -= ds;
                 else
                     closest_type[i] += ds;
             }
         }
-        auto min_act = *std::min_element(closest_type.begin(), closest_type.end());
-        auto max_act = *std::max_element(closest_type.begin(), closest_type.end());
-        for (auto& v : closest_type)
-        {
-            auto n = 2 * (v - min_act) / (max_act - min_act) - 1;
-            v = n;
-        }
         
         return closest_type;
     }
@@ -239,24 +269,41 @@ namespace assign3
             case debug_type::DATA_POINT:
             {
                 std::vector<blt::vec2> data_positions;
+                std::vector<blt::vec2> neuron_positions;
                 for (const auto& [i, v] : blt::enumerate(file.data_points))
                 {
                     auto pos = som->get_topological_position(v.bins) * 120 + blt::vec2{370, 145};
-                    auto color = blt::make_color(1,1,1);
+                    data_positions.push_back(pos);
-                    float z_index = 1;
+                    auto color = blt::make_color(1, 1, 1);
+                    float z_index = 2;
+                    if (draw_colors)
+                    {
+                        if (v.is_bad)
+                            color = blt::make_color(1, 0, 0);
+                        else
+                            color = blt::make_color(0, 1, 0);
+                    }
                     if (i == static_cast<blt::size_t>(selected_data_point))
+                    {
+                        if (!draw_colors)
                         {
                             color = blt::make_color(1, 0, 1);
-                        z_index = 2;
+                            z_index = 3;
                         }
-                    br2d.drawRectangleInternal(color, blt::gfx::rectangle2d_t{pos, blt::vec2{8,8}}, z_index);
+                    }
+                    br2d.drawRectangleInternal(color, blt::gfx::rectangle2d_t{pos, blt::vec2{8, 8}}, z_index);
                 }
                 
                 const auto& data_point = file.data_points[selected_data_point];
-                auto closest_type = get_neuron_activations(file);
+                auto closest_type = normalize_data(get_neuron_activations(file));
                 draw_som(neuron_render_info_t{}.set_base_pos({370, 145}).set_neuron_scale(120).set_neuron_padding({0, 0}),
-                         [this, &data_point, &closest_type](render_data_t context) {
+                         [this, &neuron_positions, &data_point, &closest_type](render_data_t context) {
-                             auto& text = fr2d.render_text(std::to_string(context.neuron.dist(data_point.bins)), 18).setColor(0.2, 0.2, 0.8);
+                             auto half = som->find_closest_neighbour_distance(context.index) / at_distance_measurement;
+                             auto scale = topology_function->scale(half, requested_activation);
+                             auto ds = topology_function->call(context.neuron.dist(data_point.bins), scale);
+                             
+                             neuron_positions.push_back(context.neuron_padded);
+                             auto& text = fr2d.render_text(std::to_string(ds), 18).setColor(0.2, 0.2, 0.8);
                              auto text_width = text.getAssociatedText().getTextWidth();
                              auto text_height = text.getAssociatedText().getTextHeight();
                              text.setPosition(context.neuron_padded - blt::vec2{text_width / 2.0f, text_height / 2.0f}).setZIndex(1);
@@ -264,6 +311,12 @@ namespace assign3
                              auto type = closest_type[context.index];
                              return type >= 0 ? blt::make_color(0, type, 0) : blt::make_color(-type, 0, 0);
                          });
+                
+                if (draw_data_lines)
+                {
+                    for (const auto& neuron : neuron_positions)
+                        br2d.drawLineInternal(blt::make_color(1, 1, 0), blt::gfx::line2d_t{neuron, data_positions[selected_data_point]}, 1);
+                }
             }
                 break;
             case debug_type::DISTANCE:
@@ -273,4 +326,17 @@ namespace assign3
                 break;
         }
     }
+    
+    std::vector<float> renderer_t::normalize_data(const std::vector<float>& data)
+    {
+        std::vector<float> return_data;
+        auto min_act = *std::min_element(data.begin(), data.end());
+        auto max_act = *std::max_element(data.begin(), data.end());
+        for (auto& v : data)
+        {
+            auto n = 2 * (v - min_act) / (max_act - min_act) - 1;
+            return_data.push_back(n);
+        }
+        return return_data;
+    }
 }

src/neuron.cpp

@@ -56,9 +56,8 @@ namespace assign3
     }
     
     // distance between two neurons, in 2d
-    Scalar neuron_t::distance(const neuron_t& n1, const neuron_t& n2)
+    Scalar neuron_t::distance(distance_function_t* dist_func, const neuron_t& n1, const neuron_t& n2)
     {
-        euclidean_distance_function_t dist_func;
+        return dist_func->distance({n1.get_x(), n1.get_y()}, {n2.get_x(), n2.get_y()});
-        return dist_func.distance({n1.get_x(), n1.get_y()}, {n2.get_x(), n2.get_y()});
     }
 }

src/som.cpp

@@ -57,7 +57,8 @@ namespace assign3
             {
                 if (i == v0_idx)
                     continue;
-                auto dist = basis_func->call(neuron_t::distance(v0, n), time_ratio * scale);
+                toroidal_euclidean_distance_function_t dist_func{static_cast<blt::i32>(array.get_width()), static_cast<blt::i32>(array.get_height())};
+                auto dist = basis_func->call(neuron_t::distance(&dist_func, v0, n), time_ratio * scale);
                 n.update(current_data.bins, dist, eta);
             }
         }
@@ -83,11 +84,12 @@ namespace assign3
     
     Scalar som_t::find_closest_neighbour_distance(blt::size_t v0)
     {
+        toroidal_euclidean_distance_function_t dist_func{static_cast<blt::i32>(array.get_width()), static_cast<blt::i32>(array.get_height())};
         Scalar distance_min = std::numeric_limits<Scalar>::max();
         for (const auto& [i, n] : blt::enumerate(array.get_map()))
         {
             if (i != v0)
-                distance_min = std::min(distance_min, neuron_t::distance(array.get_map()[v0], n));
+                distance_min = std::min(distance_min, neuron_t::distance(&dist_func, array.get_map()[v0], n));
         }
         return distance_min;
     }