Fix xlsx library finding

Reviewed-on: #10

.clang-tidy

@@ -0,0 +1,16 @@
+---
+Checks: '-*,
+         clang-*,
+         bugprone-*,
+         cppcoreguidelines-*,
+         modernize-*,
+         performance-*,
+         -cppcoreguidelines-pro-type-vararg,
+         -modernize-use-trailing-return-type,
+         -bugprone-exception-escape'
+
+HeaderFilterRegex: 'src/*'
+AnalyzeTemporaryDtors: false
+WarningsAsErrors: ''
+FormatStyle: file
+...

.clang-uml

@@ -0,0 +1,31 @@
+compilation_database_dir: build
+output_directory: puml
+diagrams:
+  BayesNet:
+    type: class
+    glob:
+      - src/BayesNet/*.cc
+      - src/Platform/*.cc
+    using_namespace: bayesnet
+    include:
+      namespaces:
+        - bayesnet
+        - platform
+    plantuml:
+      after:
+        - "note left of {{ alias(\"MyProjectMain\") }}: Main class of myproject library."
+  sequence:
+    type: sequence
+    glob:
+      - src/Platform/main.cc
+    combine_free_functions_into_file_participants: true
+    using_namespace:
+      - std
+      - bayesnet
+      - platform
+    include:
+      paths:
+        - src/BayesNet
+        - src/Platform
+    start_from:
+      - function: main(int,const char **)

.gitignore

@@ -31,7 +31,11 @@
 *.exe
 *.out
 *.app
-build/
+build/**
+build_debug/**
+build_release/**
 *.dSYM/**
 cmake-build*/**
 .idea
+puml/**
+.vscode/settings.json

.gitmodules

@@ -0,0 +1,15 @@
+[submodule "lib/mdlp"]
+	path = lib/mdlp
+	url = https://github.com/rmontanana/mdlp
+[submodule "lib/catch2"]
+	path = lib/catch2
+	url = https://github.com/catchorg/Catch2.git
+[submodule "lib/argparse"]
+	path = lib/argparse
+	url = https://github.com/p-ranav/argparse
+[submodule "lib/json"]
+	path = lib/json
+	url = https://github.com/nlohmann/json.git
+[submodule "lib/libxlsxwriter"]
+	path = lib/libxlsxwriter
+	url = https://github.com/jmcnamara/libxlsxwriter.git

.vscode/c_cpp_properties.json

@@ -0,0 +1,18 @@
+{
+    "configurations": [
+        {
+            "name": "Mac",
+            "includePath": [
+                "${workspaceFolder}/**"
+            ],
+            "defines": [],
+            "macFrameworkPath": [
+                "/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/System/Library/Frameworks"
+            ],
+            "cStandard": "c17",
+            "cppStandard": "c++17",
+            "compileCommands": "${workspaceFolder}/cmake-build-release/compile_commands.json"
+        }
+    ],
+    "version": 4
+}

.vscode/launch.json

@@ -4,22 +4,81 @@
         {
             "type": "lldb",
             "request": "launch",
-            "name": "bayesnet",
-            "program": "${workspaceFolder}/build/sample/main",
+            "name": "sample",
+            "program": "${workspaceFolder}/build/sample/BayesNetSample",
             "args": [
-                "-f",
-                "iris"
+                "-d",
+                "iris",
+                "-m",
+                "TANLd",
+                "-s",
+                "271",
+                "-p",
+                "/Users/rmontanana/Code/discretizbench/datasets/",
             ],
-            "cwd": "${workspaceFolder}",
-            "preLaunchTask": "CMake: build"
+            //"cwd": "${workspaceFolder}/build/sample/",
         },
         {
             "type": "lldb",
             "request": "launch",
-            "name": "aout",
-            "program": "${workspaceFolder}/a.out",
+            "name": "experiment",
+            "program": "${workspaceFolder}/build/src/Platform/b_main",
+            "args": [
+                "-m",
+                "TAN",
+                "--stratified",
+                "-d",
+                "zoo",
+                "--discretize"
+                // "--hyperparameters",
+                // "{\"repeatSparent\": true, \"maxModels\": 12}"
+            ],
+            "cwd": "/Users/rmontanana/Code/odtebench",
+        },
+        {
+            "type": "lldb",
+            "request": "launch",
+            "name": "best",
+            "program": "${workspaceFolder}/build/src/Platform/b_best",
+            "args": [
+                "-m",
+                "BoostAODE",
+                "-s",
+                "accuracy",
+                "--build",
+            ],
+            "cwd": "/Users/rmontanana/Code/discretizbench",
+        },
+        {
+            "type": "lldb",
+            "request": "launch",
+            "name": "manage",
+            "program": "${workspaceFolder}/build/src/Platform/b_manage",
+            "args": [
+                "-n",
+                "20"
+            ],
+            "cwd": "/Users/rmontanana/Code/discretizbench",
+        },
+        {
+            "type": "lldb",
+            "request": "launch",
+            "name": "list",
+            "program": "${workspaceFolder}/build/src/Platform/b_list",
             "args": [],
-            "cwd": "${workspaceFolder}"
+            //"cwd": "/Users/rmontanana/Code/discretizbench",
+            "cwd": "/home/rmontanana/Code/covbench",
+        },
+        {
+            "type": "lldb",
+            "request": "launch",
+            "name": "test",
+            "program": "${workspaceFolder}/build/tests/unit_tests",
+            "args": [
+                "-c=\"Metrics Test\"",
+                // "-s",
+            ],
+            "cwd": "${workspaceFolder}/build/tests",
         },
         {
             "name": "Build & debug active file",

.vscode/settings.json

@@ -1,104 +0,0 @@
-{
-    "files.associations": {
-        "*.rmd": "markdown",
-        "*.py": "python",
-        "vector": "cpp",
-        "__bit_reference": "cpp",
-        "__bits": "cpp",
-        "__config": "cpp",
-        "__debug": "cpp",
-        "__errc": "cpp",
-        "__hash_table": "cpp",
-        "__locale": "cpp",
-        "__mutex_base": "cpp",
-        "__node_handle": "cpp",
-        "__nullptr": "cpp",
-        "__split_buffer": "cpp",
-        "__string": "cpp",
-        "__threading_support": "cpp",
-        "__tuple": "cpp",
-        "array": "cpp",
-        "atomic": "cpp",
-        "bitset": "cpp",
-        "cctype": "cpp",
-        "chrono": "cpp",
-        "clocale": "cpp",
-        "cmath": "cpp",
-        "compare": "cpp",
-        "complex": "cpp",
-        "concepts": "cpp",
-        "cstdarg": "cpp",
-        "cstddef": "cpp",
-        "cstdint": "cpp",
-        "cstdio": "cpp",
-        "cstdlib": "cpp",
-        "cstring": "cpp",
-        "ctime": "cpp",
-        "cwchar": "cpp",
-        "cwctype": "cpp",
-        "exception": "cpp",
-        "initializer_list": "cpp",
-        "ios": "cpp",
-        "iosfwd": "cpp",
-        "istream": "cpp",
-        "limits": "cpp",
-        "locale": "cpp",
-        "memory": "cpp",
-        "mutex": "cpp",
-        "new": "cpp",
-        "optional": "cpp",
-        "ostream": "cpp",
-        "ratio": "cpp",
-        "sstream": "cpp",
-        "stdexcept": "cpp",
-        "streambuf": "cpp",
-        "string": "cpp",
-        "string_view": "cpp",
-        "system_error": "cpp",
-        "tuple": "cpp",
-        "type_traits": "cpp",
-        "typeinfo": "cpp",
-        "unordered_map": "cpp",
-        "variant": "cpp",
-        "algorithm": "cpp",
-        "iostream": "cpp",
-        "iomanip": "cpp",
-        "numeric": "cpp",
-        "set": "cpp",
-        "__tree": "cpp",
-        "deque": "cpp",
-        "list": "cpp",
-        "map": "cpp",
-        "unordered_set": "cpp",
-        "any": "cpp",
-        "condition_variable": "cpp",
-        "forward_list": "cpp",
-        "fstream": "cpp",
-        "stack": "cpp",
-        "thread": "cpp",
-        "__memory": "cpp",
-        "filesystem": "cpp",
-        "*.toml": "toml",
-        "utility": "cpp",
-        "__verbose_abort": "cpp",
-        "bit": "cpp",
-        "random": "cpp",
-        "*.tcc": "cpp",
-        "functional": "cpp",
-        "iterator": "cpp",
-        "memory_resource": "cpp",
-        "format": "cpp",
-        "valarray": "cpp",
-        "regex": "cpp",
-        "span": "cpp",
-        "cfenv": "cpp",
-        "cinttypes": "cpp",
-        "csetjmp": "cpp",
-        "future": "cpp",
-        "queue": "cpp",
-        "typeindex": "cpp",
-        "shared_mutex": "cpp"
-    },
-    "cmake.configureOnOpen": false,
-    "C_Cpp.default.configurationProvider": "ms-vscode.cmake-tools"
-}

.vscode/tasks.json

@@ -32,6 +32,29 @@
             ],
             "group": "build",
             "detail": "Task generated by Debugger."
+        },
+        {
+            "type": "cppbuild",
+            "label": "C/C++: g++ build active file",
+            "command": "/usr/bin/g++",
+            "args": [
+                "-fdiagnostics-color=always",
+                "-g",
+                "${file}",
+                "-o",
+                "${fileDirname}/${fileBasenameNoExtension}"
+            ],
+            "options": {
+                "cwd": "${fileDirname}"
+            },
+            "problemMatcher": [
+                "$gcc"
+            ],
+            "group": {
+                "kind": "build",
+                "isDefault": true
+            },
+            "detail": "Task generated by Debugger."
         }
     ]
 }

CMakeLists.txt

@@ -1,16 +1,20 @@
 cmake_minimum_required(VERSION 3.20)
 
 project(BayesNet
-  VERSION 0.1.0
+  VERSION 0.2.0
   DESCRIPTION "Bayesian Network and basic classifiers Library."
   HOMEPAGE_URL "https://github.com/rmontanana/bayesnet"
   LANGUAGES CXX
 )
 
+if (CODE_COVERAGE AND NOT ENABLE_TESTING)
+  MESSAGE(FATAL_ERROR "Code coverage requires testing enabled")
+endif (CODE_COVERAGE AND NOT ENABLE_TESTING)
+
 find_package(Torch REQUIRED)
 
 if (POLICY CMP0135)
-    cmake_policy(SET CMP0135 NEW)
+  cmake_policy(SET CMP0135 NEW)
 endif ()
 
 # Global CMake variables
@@ -24,36 +28,66 @@ set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${TORCH_CXX_FLAGS}")
 # Options
 # -------
 option(ENABLE_CLANG_TIDY "Enable to add clang tidy."              OFF)
-option(ENABLE_TESTING "Unit testing build"                        ON)
-option(CODE_COVERAGE "Collect coverage from test library"         ON)
+option(ENABLE_TESTING "Unit testing build"                        OFF)
+option(CODE_COVERAGE "Collect coverage from test library"         OFF)
 
-set(CMAKE_BUILD_TYPE "Debug")
+# Boost Library
+set(Boost_USE_STATIC_LIBS OFF) 
+set(Boost_USE_MULTITHREADED ON)  
+set(Boost_USE_STATIC_RUNTIME OFF) 
+find_package(Boost 1.66.0 REQUIRED) 
+if(Boost_FOUND)
+    message("Boost_INCLUDE_DIRS=${Boost_INCLUDE_DIRS}")
+    include_directories(${Boost_INCLUDE_DIRS}) 
+endif()
 
+SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -pthread")
 # CMakes modules
 # --------------
 set(CMAKE_MODULE_PATH ${CMAKE_CURRENT_SOURCE_DIR}/cmake/modules ${CMAKE_MODULE_PATH})
+include(AddGitSubmodule)
+if (CODE_COVERAGE)
+    enable_testing()
+    include(CodeCoverage)
+    MESSAGE("Code coverage enabled")
+    set(CMAKE_CXX_FLAGS " ${CMAKE_CXX_FLAGS} -fprofile-arcs -ftest-coverage -O0 -g")
+    SET(GCC_COVERAGE_LINK_FLAGS " ${GCC_COVERAGE_LINK_FLAGS} -lgcov --coverage")
+endif (CODE_COVERAGE)
+
+if (ENABLE_CLANG_TIDY)
+  include(StaticAnalyzers) # clang-tidy
+endif (ENABLE_CLANG_TIDY)
+
+# External libraries - dependencies of BayesNet
+# ---------------------------------------------
+# include(FetchContent)
+add_git_submodule("lib/mdlp")
+add_git_submodule("lib/argparse")
+add_git_submodule("lib/json")
+
+
+find_library(XLSXWRITER_LIB NAMES libxlsxwriter.dylib libxlsxwriter.so PATHS ${BayesNet_SOURCE_DIR}/lib/libxlsxwriter/lib)
+message("XLSXWRITER_LIB=${XLSXWRITER_LIB}")
 
 
 # Subdirectories
 # --------------
 add_subdirectory(config)
-add_subdirectory(src)
+add_subdirectory(lib/Files)
+add_subdirectory(src/BayesNet)
+add_subdirectory(src/Platform)
 add_subdirectory(sample)
 
+file(GLOB BayesNet_HEADERS CONFIGURE_DEPENDS ${BayesNet_SOURCE_DIR}/src/BayesNet/*.h ${BayesNet_SOURCE_DIR}/BayesNet/*.h)
+file(GLOB BayesNet_SOURCES CONFIGURE_DEPENDS ${BayesNet_SOURCE_DIR}/src/BayesNet/*.cc ${BayesNet_SOURCE_DIR}/src/BayesNet/*.cpp)
+file(GLOB Platform_SOURCES CONFIGURE_DEPENDS ${BayesNet_SOURCE_DIR}/src/Platform/*.cc ${BayesNet_SOURCE_DIR}/src/Platform/*.cpp)
+
 # Testing
 # -------
+
 if (ENABLE_TESTING)
   MESSAGE("Testing enabled")
-  enable_testing()
-  if (CODE_COVERAGE)
-    include(CodeCoverage)
-    MESSAGE("Code coverage enabled")
-    set(CMAKE_C_FLAGS " ${CMAKE_C_FLAGS} -fprofile-arcs -ftest-coverage")
-    set(CMAKE_CXX_FLAGS " ${CMAKE_CXX_FLAGS} -fprofile-arcs -ftest-coverage")
-    SET(GCC_COVERAGE_LINK_FLAGS " ${GCC_COVERAGE_LINK_FLAGS} -lgcov --coverage")
-  endif (CODE_COVERAGE)
-  find_package(Catch2 3 REQUIRED)
+  add_git_submodule("lib/catch2")  
   include(CTest)
-  include(Catch)
   add_subdirectory(tests)
-endif (ENABLE_TESTING)
+endif (ENABLE_TESTING)

Makefile

@@ -1,6 +1,26 @@
 SHELL := /bin/bash
 .DEFAULT_GOAL := help
-.PHONY: coverage setup help build test
+.PHONY: coverage setup help build test clean debug release
+
+f_release = build_release
+f_debug = build_debug
+app_targets = b_best b_list b_main b_manage
+test_targets = unit_tests_bayesnet unit_tests_platform
+n_procs = -j 16
+
+define ClearTests
+	@for t in $(test_targets); do \
+		if [ -f $(f_debug)/tests/$$t ]; then \
+			echo ">>> Cleaning $$t..." ; \
+			rm -f $(f_debug)/tests/$$t ; \
+		fi ; \
+	done
+	@nfiles="$(find . -name "*.gcda" -print0)" ; \
+	if test "${nfiles}" != "" ; then \
+		find . -name "*.gcda" -print0 | xargs -0 rm 2>/dev/null ;\
+	fi ; 
+endef
+
 
 setup: ## Install dependencies for tests and coverage
 	@if [ "$(shell uname)" = "Darwin" ]; then \
@@ -11,35 +31,85 @@ setup: ## Install dependencies for tests and coverage
 		pip install gcovr; \
 	fi
 
-dependency: ## Create a dependency graph diagram of the project (build/dependency.png)
-	cd build && cmake .. --graphviz=dependency.dot && dot -Tpng dependency.dot -o dependency.png
+dest ?= ${HOME}/bin
+install: ## Copy binary files to bin folder
+	@echo "Destination folder: $(dest)"
+	make buildr
+	@echo ">>> Copying files to $(dest)"
+	@cp $(f_release)/src/Platform/b_main $(dest)
+	@cp $(f_release)/src/Platform/b_list $(dest)
+	@cp $(f_release)/src/Platform/b_manage $(dest)
+	@cp $(f_release)/src/Platform/b_best $(dest)
 
-build: ## Build the project
-	@echo ">>> Building BayesNet ...";
-	@if [ -d ./build ]; then rm -rf ./build; fi
-	@mkdir build; 
-	cmake -S . -B build; \
-	cd build; \
-	make; \
-	
+dependency: ## Create a dependency graph diagram of the project (build/dependency.png)
+	@echo ">>> Creating dependency graph diagram of the project...";
+	$(MAKE) debug
+	cd $(f_debug) && cmake .. --graphviz=dependency.dot && dot -Tpng dependency.dot -o dependency.png
+
+buildd: ## Build the debug targets
+	cmake --build $(f_debug) -t $(app_targets) $(n_procs)
+
+buildr: ## Build the release targets
+	cmake --build $(f_release) -t $(app_targets) $(n_procs)
+
+clean: ## Clean the tests info
+	@echo ">>> Cleaning Debug BayesNet tests...";
+	$(call ClearTests)
 	@echo ">>> Done";
 
-test: ## Run tests
-	@echo "* Running tests...";
-	find . -name "*.gcda" -print0 | xargs -0 rm
-	@cd build; \
-	cmake --build . --target unit_tests ;
-	@cd build/tests; \
-	./unit_tests;
+clang-uml: ## Create uml class and sequence diagrams
+	clang-uml -p --add-compile-flag -I /usr/lib/gcc/x86_64-redhat-linux/8/include/
+
+debug: ## Build a debug version of the project
+	@echo ">>> Building Debug BayesNet...";
+	@if [ -d ./$(f_debug) ]; then rm -rf ./$(f_debug); fi
+	@mkdir $(f_debug); 
+	@cmake -S . -B $(f_debug) -D CMAKE_BUILD_TYPE=Debug -D ENABLE_TESTING=ON -D CODE_COVERAGE=ON
+	@echo ">>> Done";
+
+release: ## Build a Release version of the project
+	@echo ">>> Building Release BayesNet...";
+	@if [ -d ./$(f_release) ]; then rm -rf ./$(f_release); fi
+	@mkdir $(f_release); 
+	@cmake -S . -B $(f_release) -D CMAKE_BUILD_TYPE=Release
+	@echo ">>> Done";	
+
+opt = ""
+test: ## Run tests (opt="-s") to verbose output the tests, (opt="-c='Test Maximum Spanning Tree'") to run only that section
+	@echo ">>> Running BayesNet & Platform tests...";
+	@$(MAKE) clean
+	@cmake --build $(f_debug) -t $(test_targets) $(n_procs)
+	@for t in $(test_targets); do \
+		if [ -f $(f_debug)/tests/$$t ]; then \
+			cd $(f_debug)/tests ; \
+			./$$t $(opt) ; \
+		fi ; \
+	done
+	@echo ">>> Done";
+
+opt = ""
+testp: ## Run platform tests (opt="-s") to verbose output the tests, (opt="-c='Stratified Fold Test'") to run only that section
+	@echo ">>> Running Platform tests...";
+	@$(MAKE) clean
+	@cmake --build $(f_debug) --target unit_tests_platform $(n_procs)
+	@if [ -f $(f_debug)/tests/unit_tests_platform ]; then cd $(f_debug)/tests ; ./unit_tests_platform $(opt) ; fi ; 
+	@echo ">>> Done";
+
+opt = ""
+testb: ## Run BayesNet tests (opt="-s") to verbose output the tests, (opt="-c='Test Maximum Spanning Tree'") to run only that section
+	@echo ">>> Running BayesNet tests...";
+	@$(MAKE) clean
+	@cmake --build $(f_debug) --target unit_tests_bayesnet $(n_procs)
+	@if [ -f $(f_debug)/tests/unit_tests_bayesnet ]; then cd $(f_debug)/tests ; ./unit_tests_bayesnet $(opt) ; fi ; 
+	@echo ">>> Done";
 
 coverage: ## Run tests and generate coverage report (build/index.html)
-	@echo "*Building tests...";
-	find . -name "*.gcda" -print0 | xargs -0 rm
-	@cd build; \
-	cmake --build . --target unit_tests ;
-	@cd build/tests; \
-	./unit_tests;
-	gcovr ;
+	@echo ">>> Building tests with coverage...";
+	@$(MAKE) test
+	@cd $(f_debug) ; \
+	gcovr --config ../gcovr.cfg tests ;
+	@echo ">>> Done";	
+
 
 help: ## Show help message
 	@IFS=$$'\n' ; \

README.md

@@ -2,4 +2,50 @@
 
 Bayesian Network Classifier with libtorch from scratch
 
+## 0. Setup
+
+Before compiling BayesNet.
+
+### boost library
+
+[Getting Started](<https://www.boost.org/doc/libs/1_83_0/more/getting_started/index.html>)
+
+The best option is install the packages that the Linux distribution have in its repository. If this is the case:
+
+```bash
+sudo dnf install boost-devel
+```
+
+If this is not possible and the compressed packaged is installed, the following environment variable has to be set:
+
+```bash
+export BOOST_ROOT=/path/to/library/
+```
+
+### libxlswriter
+
+```bash
+cd lib/libxlsxwriter
+make
+make install DESTDIR=/home/rmontanana/Code PREFIX=
+```
+
+Environment variable has to be set:
+
+```bash
+ export LD_LIBRARY_PATH=/usr/local/lib
+ ```
+
+### Release
+
+```bash
+make release
+```
+
+### Debug & Tests
+
+```bash
+make debug
+```
+
 ## 1. Introduction

cmake/modules/AddGitSubmodule.cmake

@@ -0,0 +1,12 @@
+
+function(add_git_submodule dir)
+  find_package(Git REQUIRED) 
+
+  if(NOT EXISTS ${dir}/CMakeLists.txt)
+    message(STATUS "🚨 Adding git submodule => ${dir}")
+    execute_process(COMMAND ${GIT_EXECUTABLE}
+      submodule update --init --recursive -- ${dir}
+      WORKING_DIRECTORY ${PROJECT_SOURCE_DIR}) 
+  endif()
+  add_subdirectory(${dir})
+endfunction(add_git_submodule)

cmake/modules/StaticAnalyzers.cmake

@@ -0,0 +1,22 @@
+if(ENABLE_CLANG_TIDY)
+    find_program(CLANG_TIDY_COMMAND NAMES clang-tidy)
+
+    if(NOT CLANG_TIDY_COMMAND)
+        message(WARNING "🔴 CMake_RUN_CLANG_TIDY is ON but clang-tidy is not found!")
+        set(CMAKE_CXX_CLANG_TIDY "" CACHE STRING "" FORCE)
+    else()
+    
+        message(STATUS "🟢 CMake_RUN_CLANG_TIDY is ON")
+        set(CLANGTIDY_EXTRA_ARGS
+            "-extra-arg=-Wno-unknown-warning-option"
+        )
+        set(CMAKE_CXX_CLANG_TIDY "${CLANG_TIDY_COMMAND};-p=${CMAKE_BINARY_DIR};${CLANGTIDY_EXTRA_ARGS}" CACHE STRING "" FORCE)
+
+        add_custom_target(clang-tidy
+            COMMAND ${CMAKE_COMMAND} --build ${CMAKE_BINARY_DIR} --target ${CMAKE_PROJECT_NAME}
+            COMMAND ${CMAKE_COMMAND} --build ${CMAKE_BINARY_DIR} --target clang-tidy
+            COMMENT "Running clang-tidy..."
+        )
+       set(CMAKE_EXPORT_COMPILE_COMMANDS ON)
+    endif()
+endif(ENABLE_CLANG_TIDY)

gcovr.cfg

@@ -1,5 +1,4 @@
 filter = src/
-exclude = external/
-exclude = tests/
+exclude-directories = build/lib/
 print-summary = yes
 sort-percentage = yes

lib/Files/ArffFiles.cc

@@ -2,6 +2,7 @@
 #include <fstream>
 #include <sstream>
 #include <map>
+#include <iostream>
 
 using namespace std;
 
@@ -42,7 +43,7 @@ vector<int>& ArffFiles::getY()
     return y;
 }
 
-void ArffFiles::load(const string& fileName, bool classLast)
+void ArffFiles::loadCommon(string fileName)
 {
     ifstream file(fileName);
     if (!file.is_open()) {
@@ -74,24 +75,51 @@ void ArffFiles::load(const string& fileName, bool classLast)
     file.close();
     if (attributes.empty())
         throw invalid_argument("No attributes found");
+}
+
+void ArffFiles::load(const string& fileName, bool classLast)
+{
+    int labelIndex;
+    loadCommon(fileName);
     if (classLast) {
         className = get<0>(attributes.back());
         classType = get<1>(attributes.back());
         attributes.pop_back();
+        labelIndex = static_cast<int>(attributes.size());
     } else {
         className = get<0>(attributes.front());
         classType = get<1>(attributes.front());
         attributes.erase(attributes.begin());
+        labelIndex = 0;
     }
-    generateDataset(classLast);
-
+    generateDataset(labelIndex);
+}
+void ArffFiles::load(const string& fileName, const string& name)
+{
+    int labelIndex;
+    loadCommon(fileName);
+    bool found = false;
+    for (int i = 0; i < attributes.size(); ++i) {
+        if (attributes[i].first == name) {
+            className = get<0>(attributes[i]);
+            classType = get<1>(attributes[i]);
+            attributes.erase(attributes.begin() + i);
+            labelIndex = i;
+            found = true;
+            break;
+        }
+    }
+    if (!found) {
+        throw invalid_argument("Class name not found");
+    }
+    generateDataset(labelIndex);
 }
 
-void ArffFiles::generateDataset(bool classLast)
+void ArffFiles::generateDataset(int labelIndex)
 {
     X = vector<vector<float>>(attributes.size(), vector<float>(lines.size()));
     auto yy = vector<string>(lines.size(), "");
-    int labelIndex = classLast ? static_cast<int>(attributes.size()) : 0;
+    auto removeLines = vector<int>(); // Lines with missing values
     for (size_t i = 0; i < lines.size(); i++) {
         stringstream ss(lines[i]);
         string value;
@@ -101,10 +129,20 @@ void ArffFiles::generateDataset(bool classLast)
             if (pos++ == labelIndex) {
                 yy[i] = value;
             } else {
-                X[xIndex++][i] = stof(value);
+                if (value == "?") {
+                    X[xIndex++][i] = -1;
+                    removeLines.push_back(i);
+                } else
+                    X[xIndex++][i] = stof(value);
             }
         }
     }
+    for (auto i : removeLines) {
+        yy.erase(yy.begin() + i);
+        for (auto& x : X) {
+            x.erase(x.begin() + i);
+        }
+    }
     y = factorize(yy);
 }
 

lib/Files/ArffFiles.h

@@ -14,12 +14,12 @@ private:
     string classType;
     vector<vector<float>> X;
     vector<int> y;
-
-    void generateDataset(bool);
-
+    void generateDataset(int);
+    void loadCommon(string);
 public:
     ArffFiles();
     void load(const string&, bool = true);
+    void load(const string&, const string&);
     vector<string> getLines() const;
     unsigned long int getSize() const;
     string getClassName() const;

lib/Files/CMakeLists.txt

@@ -0,0 +1 @@
+add_library(ArffFiles ArffFiles.cc)

mac_mst.txt

@@ -0,0 +1,33 @@
+Weights matrix: 
+ 0.0000000,   0.0384968,   0.0795434,   0.1546867,  -0.0000000,   0.1788104,   0.2214721,   0.0323837,   0.0366549,  
+ 0.0384968,   0.0000000,   0.0200662,   0.0200937,  -0.0000000,   0.0637224,   0.0183005,   0.0127657,   0.0136054,  
+ 0.0795434,   0.0200662,   0.0000000,   0.0605489,  -0.0000000,   0.0894469,   0.1689408,   0.0321602,   0.0223184,  
+ 0.1546867,   0.0200937,   0.0605489,   0.0000000,  -0.0000000,   0.1150757,   0.1332292,   0.0422865,   0.0191138,  
+-0.0000000,  -0.0000000,  -0.0000000,  -0.0000000,   0.0000000,   0.0000000,   0.0000000,   0.0000000,   0.0000000,  
+ 0.1788104,   0.0637224,   0.0894469,   0.1150757,   0.0000000,   0.0000000,   0.1407102,   0.0406590,   0.0366986,  
+ 0.2214721,   0.0183005,   0.1689408,   0.1332292,   0.0000000,   0.1407102,   0.0000000,   0.0427515,   0.0349965,  
+ 0.0323837,   0.0127657,   0.0321602,   0.0422865,   0.0000000,   0.0406590,   0.0427515,   0.0000000,   0.0343376,  
+ 0.0366549,   0.0136054,   0.0223184,   0.0191138,   0.0000000,   0.0366986,   0.0349965,   0.0343376,   0.0000000,  
+Edge : Weight
+0 - 6 : 0.2214721
+0 - 5 : 0.1788104
+2 - 6 : 0.1689408
+0 - 3 : 0.1546867
+1 - 5 : 0.0637224
+6 - 7 : 0.0427515
+5 - 8 : 0.0366986
+4 - 5 : 0.0000000
+-------------------------------------------------------------------------------
+Metrics Test
+  Test Maximum Spanning Tree
+-------------------------------------------------------------------------------
+/Users/rmontanana/Code/BayesNet/tests/TestBayesMetrics.cc:58
+...............................................................................
+
+/Users/rmontanana/Code/BayesNet/tests/TestBayesMetrics.cc:69: PASSED:
+  REQUIRE( result == resultsMST.at(file_name) )
+with expansion:
+  (0, 6) (0, 5) (0, 3) (5, 1) (5, 8) (5, 4) (6, 2) (6, 7)
+  ==
+  (0, 6) (0, 5) (0, 3) (5, 1) (5, 8) (5, 4) (6, 2) (6, 7)
+

sample/CMakeLists.txt

@@ -1,4 +1,8 @@
-include_directories(${BayesNet_SOURCE_DIR}/src)
-link_directories(${MyProject_SOURCE_DIR}/src)
-add_executable(main main.cc ArffFiles.cc CPPFImdlp.cpp Metrics.cpp)
-target_link_libraries(main BayesNet "${TORCH_LIBRARIES}")
+include_directories(${BayesNet_SOURCE_DIR}/src/Platform)
+include_directories(${BayesNet_SOURCE_DIR}/src/BayesNet)
+include_directories(${BayesNet_SOURCE_DIR}/lib/Files)
+include_directories(${BayesNet_SOURCE_DIR}/lib/mdlp)
+include_directories(${BayesNet_SOURCE_DIR}/lib/argparse/include)
+include_directories(${BayesNet_SOURCE_DIR}/lib/json/include)
+add_executable(BayesNetSample sample.cc ${BayesNet_SOURCE_DIR}/src/Platform/Folding.cc ${BayesNet_SOURCE_DIR}/src/Platform/Models.cc) 
+target_link_libraries(BayesNetSample BayesNet ArffFiles mdlp "${TORCH_LIBRARIES}")

sample/CPPFImdlp.cpp

@@ -1,221 +0,0 @@
-#include <numeric>
-#include <algorithm>
-#include <set>
-#include <cmath>
-#include "CPPFImdlp.h"
-#include "Metrics.h"
-
-namespace mdlp {
-
-    CPPFImdlp::CPPFImdlp(size_t min_length_, int max_depth_, float proposed) : min_length(min_length_),
-        max_depth(max_depth_),
-        proposed_cuts(proposed)
-    {
-    }
-
-    CPPFImdlp::CPPFImdlp() = default;
-
-    CPPFImdlp::~CPPFImdlp() = default;
-
-    size_t CPPFImdlp::compute_max_num_cut_points() const
-    {
-        // Set the actual maximum number of cut points as a number or as a percentage of the number of samples
-        if (proposed_cuts == 0) {
-            return numeric_limits<size_t>::max();
-        }
-        if (proposed_cuts < 0 || proposed_cuts > static_cast<float>(X.size())) {
-            throw invalid_argument("wrong proposed num_cuts value");
-        }
-        if (proposed_cuts < 1)
-            return static_cast<size_t>(round(static_cast<float>(X.size()) * proposed_cuts));
-        return static_cast<size_t>(proposed_cuts);
-    }
-
-    void CPPFImdlp::fit(samples_t& X_, labels_t& y_)
-    {
-        X = X_;
-        y = y_;
-        num_cut_points = compute_max_num_cut_points();
-        depth = 0;
-        discretizedData.clear();
-        cutPoints.clear();
-        if (X.size() != y.size()) {
-            throw invalid_argument("X and y must have the same size");
-        }
-        if (X.empty() || y.empty()) {
-            throw invalid_argument("X and y must have at least one element");
-        }
-        if (min_length < 3) {
-            throw invalid_argument("min_length must be greater than 2");
-        }
-        if (max_depth < 1) {
-            throw invalid_argument("max_depth must be greater than 0");
-        }
-        indices = sortIndices(X_, y_);
-        metrics.setData(y, indices);
-        computeCutPoints(0, X.size(), 1);
-        sort(cutPoints.begin(), cutPoints.end());
-        if (num_cut_points > 0) {
-            // Select the best (with lower entropy) cut points
-            while (cutPoints.size() > num_cut_points) {
-                resizeCutPoints();
-            }
-        }
-    }
-
-    pair<precision_t, size_t> CPPFImdlp::valueCutPoint(size_t start, size_t cut, size_t end)
-    {
-        size_t n;
-        size_t m;
-        size_t idxPrev = cut - 1 >= start ? cut - 1 : cut;
-        size_t idxNext = cut + 1 < end ? cut + 1 : cut;
-        bool backWall; // true if duplicates reach beginning of the interval
-        precision_t previous;
-        precision_t actual;
-        precision_t next;
-        previous = X[indices[idxPrev]];
-        actual = X[indices[cut]];
-        next = X[indices[idxNext]];
-        // definition 2 of the paper => X[t-1] < X[t]
-        // get the first equal value of X in the interval
-        while (idxPrev > start && actual == previous) {
-            previous = X[indices[--idxPrev]];
-        }
-        backWall = idxPrev == start && actual == previous;
-        // get the last equal value of X in the interval
-        while (idxNext < end - 1 && actual == next) {
-            next = X[indices[++idxNext]];
-        }
-        // # of duplicates before cutpoint
-        n = cut - 1 - idxPrev;
-        // # of duplicates after cutpoint
-        m = idxNext - cut - 1;
-        // Decide which values to use
-        cut = cut + (backWall ? m + 1 : -n);
-        actual = X[indices[cut]];
-        return { (actual + previous) / 2, cut };
-    }
-
-    void CPPFImdlp::computeCutPoints(size_t start, size_t end, int depth_)
-    {
-        size_t cut;
-        pair<precision_t, size_t> result;
-        // Check if the interval length and the depth are Ok
-        if (end - start < min_length || depth_ > max_depth)
-            return;
-        depth = depth_ > depth ? depth_ : depth;
-        cut = getCandidate(start, end);
-        if (cut == numeric_limits<size_t>::max())
-            return;
-        if (mdlp(start, cut, end)) {
-            result = valueCutPoint(start, cut, end);
-            cut = result.second;
-            cutPoints.push_back(result.first);
-            computeCutPoints(start, cut, depth_ + 1);
-            computeCutPoints(cut, end, depth_ + 1);
-        }
-    }
-
-    size_t CPPFImdlp::getCandidate(size_t start, size_t end)
-    {
-        /* Definition 1: A binary discretization for A is determined by selecting the cut point TA for which
-        E(A, TA; S) is minimal amongst all the candidate cut points. */
-        size_t candidate = numeric_limits<size_t>::max();
-        size_t elements = end - start;
-        bool sameValues = true;
-        precision_t entropy_left;
-        precision_t entropy_right;
-        precision_t minEntropy;
-        // Check if all the values of the variable in the interval are the same
-        for (size_t idx = start + 1; idx < end; idx++) {
-            if (X[indices[idx]] != X[indices[start]]) {
-                sameValues = false;
-                break;
-            }
-        }
-        if (sameValues)
-            return candidate;
-        minEntropy = metrics.entropy(start, end);
-        for (size_t idx = start + 1; idx < end; idx++) {
-            // Cutpoints are always on boundaries (definition 2)
-            if (y[indices[idx]] == y[indices[idx - 1]])
-                continue;
-            entropy_left = precision_t(idx - start) / static_cast<precision_t>(elements) * metrics.entropy(start, idx);
-            entropy_right = precision_t(end - idx) / static_cast<precision_t>(elements) * metrics.entropy(idx, end);
-            if (entropy_left + entropy_right < minEntropy) {
-                minEntropy = entropy_left + entropy_right;
-                candidate = idx;
-            }
-        }
-        return candidate;
-    }
-
-    bool CPPFImdlp::mdlp(size_t start, size_t cut, size_t end)
-    {
-        int k;
-        int k1;
-        int k2;
-        precision_t ig;
-        precision_t delta;
-        precision_t ent;
-        precision_t ent1;
-        precision_t ent2;
-        auto N = precision_t(end - start);
-        k = metrics.computeNumClasses(start, end);
-        k1 = metrics.computeNumClasses(start, cut);
-        k2 = metrics.computeNumClasses(cut, end);
-        ent = metrics.entropy(start, end);
-        ent1 = metrics.entropy(start, cut);
-        ent2 = metrics.entropy(cut, end);
-        ig = metrics.informationGain(start, cut, end);
-        delta = static_cast<precision_t>(log2(pow(3, precision_t(k)) - 2) -
-            (precision_t(k) * ent - precision_t(k1) * ent1 - precision_t(k2) * ent2));
-        precision_t term = 1 / N * (log2(N - 1) + delta);
-        return ig > term;
-    }
-
-    // Argsort from https://stackoverflow.com/questions/1577475/c-sorting-and-keeping-track-of-indexes
-    indices_t CPPFImdlp::sortIndices(samples_t& X_, labels_t& y_)
-    {
-        indices_t idx(X_.size());
-        iota(idx.begin(), idx.end(), 0);
-        stable_sort(idx.begin(), idx.end(), [&X_, &y_](size_t i1, size_t i2) {
-            if (X_[i1] == X_[i2])
-                return y_[i1] < y_[i2];
-            else
-                return X_[i1] < X_[i2];
-            });
-        return idx;
-    }
-
-    void CPPFImdlp::resizeCutPoints()
-    {
-        //Compute entropy of each of the whole cutpoint set and discards the biggest value
-        precision_t maxEntropy = 0;
-        precision_t entropy;
-        size_t maxEntropyIdx = 0;
-        size_t begin = 0;
-        size_t end;
-        for (size_t idx = 0; idx < cutPoints.size(); idx++) {
-            end = begin;
-            while (X[indices[end]] < cutPoints[idx] && end < X.size())
-                end++;
-            entropy = metrics.entropy(begin, end);
-            if (entropy > maxEntropy) {
-                maxEntropy = entropy;
-                maxEntropyIdx = idx;
-            }
-            begin = end;
-        }
-        cutPoints.erase(cutPoints.begin() + static_cast<long>(maxEntropyIdx));
-    }
-    labels_t& CPPFImdlp::transform(const samples_t& data)
-    {
-        discretizedData.reserve(data.size());
-        for (const precision_t& item : data) {
-            auto upper = upper_bound(cutPoints.begin(), cutPoints.end(), item);
-            discretizedData.push_back(upper - cutPoints.begin());
-        }
-        return discretizedData;
-    }
-}

sample/CPPFImdlp.h

@@ -1,45 +0,0 @@
-#ifndef CPPFIMDLP_H
-#define CPPFIMDLP_H
-
-#include "typesFImdlp.h"
-#include "Metrics.h"
-#include <limits>
-#include <utility>
-#include <string>
-
-namespace mdlp {
-    class CPPFImdlp {
-    protected:
-        size_t min_length = 3;
-        int depth = 0;
-        int max_depth = numeric_limits<int>::max();
-        float proposed_cuts = 0;
-        indices_t indices = indices_t();
-        samples_t X = samples_t();
-        labels_t y = labels_t();
-        Metrics metrics = Metrics(y, indices);
-        cutPoints_t cutPoints;
-        size_t num_cut_points = numeric_limits<size_t>::max();
-        labels_t discretizedData = labels_t();
-
-        static indices_t sortIndices(samples_t&, labels_t&);
-
-        void computeCutPoints(size_t, size_t, int);
-        void resizeCutPoints();
-        bool mdlp(size_t, size_t, size_t);
-        size_t getCandidate(size_t, size_t);
-        size_t compute_max_num_cut_points() const;
-        pair<precision_t, size_t> valueCutPoint(size_t, size_t, size_t);
-
-    public:
-        CPPFImdlp();
-        CPPFImdlp(size_t, int, float);
-        ~CPPFImdlp();
-        void fit(samples_t&, labels_t&);
-        inline cutPoints_t getCutPoints() const { return cutPoints; };
-        labels_t& transform(const samples_t&);
-        inline int get_depth() const { return depth; };
-        static inline string version() { return "1.1.2"; };
-    };
-}
-#endif

sample/Metrics.cpp

@@ -1,78 +0,0 @@
-#include "Metrics.h"
-#include <set>
-#include <cmath>
-
-using namespace std;
-namespace mdlp {
-    Metrics::Metrics(labels_t& y_, indices_t& indices_): y(y_), indices(indices_),
-        numClasses(computeNumClasses(0, indices.size()))
-    {
-    }
-
-    int Metrics::computeNumClasses(size_t start, size_t end)
-    {
-        set<int> nClasses;
-        for (auto i = start; i < end; ++i) {
-            nClasses.insert(y[indices[i]]);
-        }
-        return static_cast<int>(nClasses.size());
-    }
-
-    void Metrics::setData(const labels_t& y_, const indices_t& indices_)
-    {
-        indices = indices_;
-        y = y_;
-        numClasses = computeNumClasses(0, indices.size());
-        entropyCache.clear();
-        igCache.clear();
-    }
-
-    precision_t Metrics::entropy(size_t start, size_t end)
-    {
-        precision_t p;
-        precision_t ventropy = 0;
-        int nElements = 0;
-        labels_t counts(numClasses + 1, 0);
-        if (end - start < 2)
-            return 0;
-        if (entropyCache.find({ start, end }) != entropyCache.end()) {
-            return entropyCache[{start, end}];
-        }
-        for (auto i = &indices[start]; i != &indices[end]; ++i) {
-            counts[y[*i]]++;
-            nElements++;
-        }
-        for (auto count : counts) {
-            if (count > 0) {
-                p = static_cast<precision_t>(count) / static_cast<precision_t>(nElements);
-                ventropy -= p * log2(p);
-            }
-        }
-        entropyCache[{start, end}] = ventropy;
-        return ventropy;
-    }
-
-    precision_t Metrics::informationGain(size_t start, size_t cut, size_t end)
-    {
-        precision_t iGain;
-        precision_t entropyInterval;
-        precision_t entropyLeft;
-        precision_t entropyRight;
-        size_t nElementsLeft = cut - start;
-        size_t nElementsRight = end - cut;
-        size_t nElements = end - start;
-        if (igCache.find(make_tuple(start, cut, end)) != igCache.end()) {
-            return igCache[make_tuple(start, cut, end)];
-        }
-        entropyInterval = entropy(start, end);
-        entropyLeft = entropy(start, cut);
-        entropyRight = entropy(cut, end);
-        iGain = entropyInterval -
-            (static_cast<precision_t>(nElementsLeft) * entropyLeft +
-                static_cast<precision_t>(nElementsRight) * entropyRight) /
-            static_cast<precision_t>(nElements);
-        igCache[make_tuple(start, cut, end)] = iGain;
-        return iGain;
-    }
-
-}

sample/Metrics.h

@@ -1,22 +0,0 @@
-#ifndef CCMETRICS_H
-#define CCMETRICS_H
-
-#include "typesFImdlp.h"
-
-namespace mdlp {
-    class Metrics {
-    protected:
-        labels_t& y;
-        indices_t& indices;
-        int numClasses;
-        cacheEnt_t entropyCache = cacheEnt_t();
-        cacheIg_t igCache = cacheIg_t();
-    public:
-        Metrics(labels_t&, indices_t&);
-        void setData(const labels_t&, const indices_t&);
-        int computeNumClasses(size_t, size_t);
-        precision_t entropy(size_t, size_t);
-        precision_t informationGain(size_t, size_t, size_t);
-    };
-}
-#endif

sample/main.cc

@@ -1,225 +0,0 @@
-#include <iostream>
-#include <string>
-#include <torch/torch.h>
-#include <thread>
-#include <getopt.h>
-#include "ArffFiles.h"
-#include "Network.h"
-#include "Metrics.hpp"
-#include "CPPFImdlp.h"
-#include "KDB.h"
-#include "SPODE.h"
-#include "AODE.h"
-#include "TAN.h"
-
-
-using namespace std;
-
-const string PATH = "data/";
-
-/* print a description of all supported options */
-void usage(const char* path)
-{
-    /* take only the last portion of the path */
-    const char* basename = strrchr(path, '/');
-    basename = basename ? basename + 1 : path;
-
-    cout << "usage: " << basename << "[OPTION]" << endl;
-    cout << "  -h, --help\t\t Print this help and exit." << endl;
-    cout
-        << "  -f, --file[=FILENAME]\t {diabetes, glass, iris, kdd_JapaneseVowels, letter, liver-disorders, mfeat-factors}."
-        << endl;
-    cout << "  -p, --path[=FILENAME]\t folder where the data files are located, default " << PATH << endl;
-    cout << "  -m, --model={AODE, KDB, SPODE, TAN}\t " << endl;
-}
-
-tuple<string, string, string> parse_arguments(int argc, char** argv)
-{
-    string file_name;
-    string model_name;
-    string path = PATH;
-    const vector<struct option> long_options = {
-            {"help",          no_argument,       nullptr, 'h'},
-            {"file",          required_argument, nullptr, 'f'},
-            {"path",          required_argument, nullptr, 'p'},
-            {"model",         required_argument, nullptr, 'm'},
-            {nullptr,         no_argument,       nullptr, 0}
-    };
-    while (true) {
-        const auto c = getopt_long(argc, argv, "hf:p:m:", long_options.data(), nullptr);
-        if (c == -1)
-            break;
-        switch (c) {
-            case 'h':
-                usage(argv[0]);
-                exit(0);
-            case 'f':
-                file_name = string(optarg);
-                break;
-            case 'm':
-                model_name = string(optarg);
-                break;
-            case 'p':
-                path = optarg;
-                if (path.back() != '/')
-                    path += '/';
-                break;
-            case '?':
-                usage(argv[0]);
-                exit(1);
-            default:
-                abort();
-        }
-    }
-    if (file_name.empty()) {
-        usage(argv[0]);
-        exit(1);
-    }
-    return make_tuple(file_name, path, model_name);
-}
-
-inline constexpr auto hash_conv(const std::string_view sv)
-{
-    unsigned long hash{ 5381 };
-    for (unsigned char c : sv) {
-        hash = ((hash << 5) + hash) ^ c;
-    }
-    return hash;
-}
-
-inline constexpr auto operator"" _sh(const char* str, size_t len)
-{
-    return hash_conv(std::string_view{ str, len });
-}
-
-pair<vector<mdlp::labels_t>, map<string, int>> discretize(vector<mdlp::samples_t>& X, mdlp::labels_t& y, vector<string> features)
-{
-    vector<mdlp::labels_t>Xd;
-    map<string, int> maxes;
-
-    auto fimdlp = mdlp::CPPFImdlp();
-    for (int i = 0; i < X.size(); i++) {
-        fimdlp.fit(X[i], y);
-        mdlp::labels_t& xd = fimdlp.transform(X[i]);
-        maxes[features[i]] = *max_element(xd.begin(), xd.end()) + 1;
-        Xd.push_back(xd);
-    }
-    return { Xd, maxes };
-}
-
-bool file_exists(const std::string& name)
-{
-    if (FILE* file = fopen(name.c_str(), "r")) {
-        fclose(file);
-        return true;
-    } else {
-        return false;
-    }
-}
-
-tuple<string, string, string> get_options(int argc, char** argv)
-{
-    map<string, bool> datasets = {
-            {"diabetes",           true},
-            {"ecoli",              true},
-            {"glass",              true},
-            {"iris",               true},
-            {"kdd_JapaneseVowels", false},
-            {"letter",             true},
-            {"liver-disorders",    true},
-            {"mfeat-factors",      true},
-    };
-    vector <string> models = { "AODE", "KDB", "SPODE", "TAN" };
-    string file_name;
-    string path;
-    string model_name;
-    tie(file_name, path, model_name) = parse_arguments(argc, argv);
-    if (datasets.find(file_name) == datasets.end()) {
-        cout << "Invalid file name: " << file_name << endl;
-        usage(argv[0]);
-        exit(1);
-    }
-    if (!file_exists(path + file_name + ".arff")) {
-        cout << "Data File " << path + file_name + ".arff" << " does not exist" << endl;
-        usage(argv[0]);
-        exit(1);
-    }
-    if (find(models.begin(), models.end(), model_name) == models.end()) {
-        cout << "Invalid model name: " << model_name << endl;
-        usage(argv[0]);
-        exit(1);
-    }
-    return { file_name, path, model_name };
-}
-
-int main(int argc, char** argv)
-{
-    string file_name, path, model_name;
-    tie(file_name, path, model_name) = get_options(argc, argv);
-    auto handler = ArffFiles();
-    handler.load(path + file_name + ".arff");
-    // Get Dataset X, y
-    vector<mdlp::samples_t>& X = handler.getX();
-    mdlp::labels_t& y = handler.getY();
-    // Get className & Features
-    auto className = handler.getClassName();
-    vector<string> features;
-    for (auto feature : handler.getAttributes()) {
-        features.push_back(feature.first);
-    }
-    // Discretize Dataset
-    vector<mdlp::labels_t> Xd;
-    map<string, int> maxes;
-    tie(Xd, maxes) = discretize(X, y, features);
-    maxes[className] = *max_element(y.begin(), y.end()) + 1;
-    map<string, vector<int>> states;
-    for (auto feature : features) {
-        states[feature] = vector<int>(maxes[feature]);
-    }
-    states[className] = vector<int>(
-        maxes[className]);
-    double score;
-    vector<string> lines;
-    vector<string> graph;
-    auto kdb = bayesnet::KDB(2);
-    auto aode = bayesnet::AODE();
-    auto spode = bayesnet::SPODE(2);
-    auto tan = bayesnet::TAN();
-    switch (hash_conv(model_name)) {
-        case "AODE"_sh:
-            aode.fit(Xd, y, features, className, states);
-            lines = aode.show();
-            score = aode.score(Xd, y);
-            graph = aode.graph();
-            break;
-        case "KDB"_sh:
-            kdb.fit(Xd, y, features, className, states);
-            lines = kdb.show();
-            score = kdb.score(Xd, y);
-            graph = kdb.graph();
-            break;
-        case "SPODE"_sh:
-            spode.fit(Xd, y, features, className, states);
-            lines = spode.show();
-            score = spode.score(Xd, y);
-            graph = spode.graph();
-            break;
-        case "TAN"_sh:
-            tan.fit(Xd, y, features, className, states);
-            lines = tan.show();
-            score = tan.score(Xd, y);
-            graph = tan.graph();
-            break;
-    }
-    for (auto line : lines) {
-        cout << line << endl;
-    }
-    cout << "Score: " << score << endl;
-    auto dot_file = model_name + "_" + file_name;
-    ofstream file(dot_file + ".dot");
-    file << graph;
-    file.close();
-    cout << "Graph saved in " << model_name << "_" << file_name << ".dot" << endl;
-    cout << "dot -Tpng -o " + dot_file + ".png " + dot_file + ".dot " << endl;
-    return 0;
-}

sample/sample.cc

@@ -0,0 +1,237 @@
+#include <iostream>
+#include <torch/torch.h>
+#include <string>
+#include <map>
+#include <argparse/argparse.hpp>
+#include <nlohmann/json.hpp>
+#include "ArffFiles.h"v
+#include "BayesMetrics.h"
+#include "CPPFImdlp.h"
+#include "Folding.h"
+#include "Models.h"
+#include "modelRegister.h"
+#include <fstream>
+
+using namespace std;
+
+const string PATH = "../../data/";
+
+pair<vector<mdlp::labels_t>, map<string, int>> discretize(vector<mdlp::samples_t>& X, mdlp::labels_t& y, vector<string> features)
+{
+    vector<mdlp::labels_t>Xd;
+    map<string, int> maxes;
+
+    auto fimdlp = mdlp::CPPFImdlp();
+    for (int i = 0; i < X.size(); i++) {
+        fimdlp.fit(X[i], y);
+        mdlp::labels_t& xd = fimdlp.transform(X[i]);
+        maxes[features[i]] = *max_element(xd.begin(), xd.end()) + 1;
+        Xd.push_back(xd);
+    }
+    return { Xd, maxes };
+}
+
+bool file_exists(const std::string& name)
+{
+    if (FILE* file = fopen(name.c_str(), "r")) {
+        fclose(file);
+        return true;
+    } else {
+        return false;
+    }
+}
+pair<vector<vector<int>>, vector<int>> extract_indices(vector<int> indices, vector<vector<int>> X, vector<int> y)
+{
+    vector<vector<int>> Xr; // nxm
+    vector<int> yr;
+    for (int col = 0; col < X.size(); ++col) {
+        Xr.push_back(vector<int>());
+    }
+    for (auto index : indices) {
+        for (int col = 0; col < X.size(); ++col) {
+            Xr[col].push_back(X[col][index]);
+        }
+        yr.push_back(y[index]);
+    }
+    return { Xr, yr };
+}
+
+int main(int argc, char** argv)
+{
+    map<string, bool> datasets = {
+            {"diabetes",           true},
+            {"ecoli",              true},
+            {"glass",              true},
+            {"iris",               true},
+            {"kdd_JapaneseVowels", false},
+            {"letter",             true},
+            {"liver-disorders",    true},
+            {"mfeat-factors",      true},
+    };
+    auto valid_datasets = vector<string>();
+    transform(datasets.begin(), datasets.end(), back_inserter(valid_datasets),
+        [](const pair<string, bool>& pair) { return pair.first; });
+    argparse::ArgumentParser program("BayesNetSample");
+    program.add_argument("-d", "--dataset")
+        .help("Dataset file name")
+        .action([valid_datasets](const std::string& value) {
+        if (find(valid_datasets.begin(), valid_datasets.end(), value) != valid_datasets.end()) {
+            return value;
+        }
+        throw runtime_error("file must be one of {diabetes, ecoli, glass, iris, kdd_JapaneseVowels, letter, liver-disorders, mfeat-factors}");
+            }
+    );
+    program.add_argument("-p", "--path")
+        .help(" folder where the data files are located, default")
+        .default_value(string{ PATH }
+    );
+    program.add_argument("-m", "--model")
+        .help("Model to use " + platform::Models::instance()->toString())
+        .action([](const std::string& value) {
+        static const vector<string> choices = platform::Models::instance()->getNames();
+        if (find(choices.begin(), choices.end(), value) != choices.end()) {
+            return value;
+        }
+        throw runtime_error("Model must be one of " + platform::Models::instance()->toString());
+            }
+    );
+    program.add_argument("--discretize").help("Discretize input dataset").default_value(false).implicit_value(true);
+    program.add_argument("--dumpcpt").help("Dump CPT Tables").default_value(false).implicit_value(true);
+    program.add_argument("--stratified").help("If Stratified KFold is to be done").default_value(false).implicit_value(true);
+    program.add_argument("--tensors").help("Use tensors to store samples").default_value(false).implicit_value(true);
+    program.add_argument("-f", "--folds").help("Number of folds").default_value(5).scan<'i', int>().action([](const string& value) {
+        try {
+            auto k = stoi(value);
+            if (k < 2) {
+                throw runtime_error("Number of folds must be greater than 1");
+            }
+            return k;
+        }
+        catch (const runtime_error& err) {
+            throw runtime_error(err.what());
+        }
+        catch (...) {
+            throw runtime_error("Number of folds must be an integer");
+        }});
+    program.add_argument("-s", "--seed").help("Random seed").default_value(-1).scan<'i', int>();
+    bool class_last, stratified, tensors, dump_cpt;
+    string model_name, file_name, path, complete_file_name;
+    int nFolds, seed;
+    try {
+        program.parse_args(argc, argv);
+        file_name = program.get<string>("dataset");
+        path = program.get<string>("path");
+        model_name = program.get<string>("model");
+        complete_file_name = path + file_name + ".arff";
+        stratified = program.get<bool>("stratified");
+        tensors = program.get<bool>("tensors");
+        nFolds = program.get<int>("folds");
+        seed = program.get<int>("seed");
+        dump_cpt = program.get<bool>("dumpcpt");
+        class_last = datasets[file_name];
+        if (!file_exists(complete_file_name)) {
+            throw runtime_error("Data File " + path + file_name + ".arff" + " does not exist");
+        }
+    }
+    catch (const exception& err) {
+        cerr << err.what() << endl;
+        cerr << program;
+        exit(1);
+    }
+
+    /*
+    * Begin Processing
+    */
+    auto handler = ArffFiles();
+    handler.load(complete_file_name, class_last);
+    // Get Dataset X, y
+    vector<mdlp::samples_t>& X = handler.getX();
+    mdlp::labels_t& y = handler.getY();
+    // Get className & Features
+    auto className = handler.getClassName();
+    vector<string> features;
+    auto attributes = handler.getAttributes();
+    transform(attributes.begin(), attributes.end(), back_inserter(features),
+        [](const pair<string, string>& item) { return item.first; });
+    // Discretize Dataset
+    auto [Xd, maxes] = discretize(X, y, features);
+    maxes[className] = *max_element(y.begin(), y.end()) + 1;
+    map<string, vector<int>> states;
+    for (auto feature : features) {
+        states[feature] = vector<int>(maxes[feature]);
+    }
+    states[className] = vector<int>(maxes[className]);
+    auto clf = platform::Models::instance()->create(model_name);
+    clf->fit(Xd, y, features, className, states);
+    if (dump_cpt) {
+        cout << "--- CPT Tables ---" << endl;
+        clf->dump_cpt();
+    }
+    auto lines = clf->show();
+    for (auto line : lines) {
+        cout << line << endl;
+    }
+    cout << "--- Topological Order ---" << endl;
+    auto order = clf->topological_order();
+    for (auto name : order) {
+        cout << name << ", ";
+    }
+    cout << "end." << endl;
+    auto score = clf->score(Xd, y);
+    cout << "Score: " << score << endl;
+    auto graph = clf->graph();
+    auto dot_file = model_name + "_" + file_name;
+    ofstream file(dot_file + ".dot");
+    file << graph;
+    file.close();
+    cout << "Graph saved in " << model_name << "_" << file_name << ".dot" << endl;
+    cout << "dot -Tpng -o " + dot_file + ".png " + dot_file + ".dot " << endl;
+    string stratified_string = stratified ? " Stratified" : "";
+    cout << nFolds << " Folds" << stratified_string << " Cross validation" << endl;
+    cout << "==========================================" << endl;
+    torch::Tensor Xt = torch::zeros({ static_cast<int>(Xd.size()), static_cast<int>(Xd[0].size()) }, torch::kInt32);
+    torch::Tensor yt = torch::tensor(y, torch::kInt32);
+    for (int i = 0; i < features.size(); ++i) {
+        Xt.index_put_({ i, "..." }, torch::tensor(Xd[i], torch::kInt32));
+    }
+    float total_score = 0, total_score_train = 0, score_train, score_test;
+    platform::Fold* fold;
+    if (stratified)
+        fold = new platform::StratifiedKFold(nFolds, y, seed);
+    else
+        fold = new platform::KFold(nFolds, y.size(), seed);
+    for (auto i = 0; i < nFolds; ++i) {
+        auto [train, test] = fold->getFold(i);
+        cout << "Fold: " << i + 1 << endl;
+        if (tensors) {
+            auto ttrain = torch::tensor(train, torch::kInt64);
+            auto ttest = torch::tensor(test, torch::kInt64);
+            torch::Tensor Xtraint = torch::index_select(Xt, 1, ttrain);
+            torch::Tensor ytraint = yt.index({ ttrain });
+            torch::Tensor Xtestt = torch::index_select(Xt, 1, ttest);
+            torch::Tensor ytestt = yt.index({ ttest });
+            clf->fit(Xtraint, ytraint, features, className, states);
+            auto temp = clf->predict(Xtraint);
+            score_train = clf->score(Xtraint, ytraint);
+            score_test = clf->score(Xtestt, ytestt);
+        } else {
+            auto [Xtrain, ytrain] = extract_indices(train, Xd, y);
+            auto [Xtest, ytest] = extract_indices(test, Xd, y);
+            clf->fit(Xtrain, ytrain, features, className, states);
+            score_train = clf->score(Xtrain, ytrain);
+            score_test = clf->score(Xtest, ytest);
+        }
+        if (dump_cpt) {
+            cout << "--- CPT Tables ---" << endl;
+            clf->dump_cpt();
+        }
+        total_score_train += score_train;
+        total_score += score_test;
+        cout << "Score Train: " << score_train << endl;
+        cout << "Score Test : " << score_test << endl;
+        cout << "-------------------------------------------------------------------------------" << endl;
+    }
+    cout << "**********************************************************************************" << endl;
+    cout << "Average Score Train: " << total_score_train / nFolds << endl;
+    cout << "Average Score Test : " << total_score / nFolds << endl;return 0;
+}

sample/typesFImdlp.h

@@ -1,18 +0,0 @@
-#ifndef TYPES_H
-#define TYPES_H
-
-#include <vector>
-#include <map>
-#include <stdexcept>
-
-using namespace std;
-namespace mdlp {
-    typedef float precision_t;
-    typedef vector<precision_t> samples_t;
-    typedef vector<int> labels_t;
-    typedef vector<size_t> indices_t;
-    typedef vector<precision_t> cutPoints_t;
-    typedef map<pair<int, int>, precision_t> cacheEnt_t;
-    typedef map<tuple<int, int, int>, precision_t> cacheIg_t;
-}
-#endif

src/BaseClassifier.cc

@@ -1,119 +0,0 @@
-#include "BaseClassifier.h"
-#include "utils.h"
-
-namespace bayesnet {
-    using namespace std;
-    using namespace torch;
-
-    BaseClassifier::BaseClassifier(Network model) : model(model), m(0), n(0), metrics(Metrics()), fitted(false) {}
-    BaseClassifier& BaseClassifier::build(vector<string>& features, string className, map<string, vector<int>>& states)
-    {
-
-        dataset = torch::cat({ X, y.view({y.size(0), 1}) }, 1);
-        this->features = features;
-        this->className = className;
-        this->states = states;
-        checkFitParameters();
-        auto n_classes = states[className].size();
-        metrics = Metrics(dataset, features, className, n_classes);
-        train();
-        model.fit(Xv, yv, features, className);
-        fitted = true;
-        return *this;
-    }
-    BaseClassifier& BaseClassifier::fit(vector<vector<int>>& X, vector<int>& y, vector<string>& features, string className, map<string, vector<int>>& states)
-    {
-        this->X = torch::zeros({ static_cast<int64_t>(X[0].size()), static_cast<int64_t>(X.size()) }, kInt64);
-        Xv = X;
-        for (int i = 0; i < X.size(); ++i) {
-            this->X.index_put_({ "...", i }, torch::tensor(X[i], kInt64));
-        }
-        this->y = torch::tensor(y, kInt64);
-        yv = y;
-        return build(features, className, states);
-    }
-    void BaseClassifier::checkFitParameters()
-    {
-        auto sizes = X.sizes();
-        m = sizes[0];
-        n = sizes[1];
-        if (m != y.size(0)) {
-            throw invalid_argument("X and y must have the same number of samples");
-        }
-        if (n != features.size()) {
-            throw invalid_argument("X and features must have the same number of features");
-        }
-        if (states.find(className) == states.end()) {
-            throw invalid_argument("className not found in states");
-        }
-        for (auto feature : features) {
-            if (states.find(feature) == states.end()) {
-                throw invalid_argument("feature [" + feature + "] not found in states");
-            }
-        }
-    }
-
-    Tensor BaseClassifier::predict(Tensor& X)
-    {
-        if (!fitted) {
-            throw logic_error("Classifier has not been fitted");
-        }
-        auto m_ = X.size(0);
-        auto n_ = X.size(1);
-        vector<vector<int>> Xd(n_, vector<int>(m_, 0));
-        for (auto i = 0; i < n_; i++) {
-            auto temp = X.index({ "...", i });
-            Xd[i] = vector<int>(temp.data_ptr<int>(), temp.data_ptr<int>() + m_);
-        }
-        auto yp = model.predict(Xd);
-        auto ypred = torch::tensor(yp, torch::kInt64);
-        return ypred;
-    }
-    vector<int> BaseClassifier::predict(vector<vector<int>>& X)
-    {
-        if (!fitted) {
-            throw logic_error("Classifier has not been fitted");
-        }
-        auto m_ = X[0].size();
-        auto n_ = X.size();
-        vector<vector<int>> Xd(n_, vector<int>(m_, 0));
-        for (auto i = 0; i < n_; i++) {
-            Xd[i] = vector<int>(X[i].begin(), X[i].end());
-        }
-        auto yp = model.predict(Xd);
-        return yp;
-    }
-    float BaseClassifier::score(Tensor& X, Tensor& y)
-    {
-        if (!fitted) {
-            throw logic_error("Classifier has not been fitted");
-        }
-        Tensor y_pred = predict(X);
-        return (y_pred == y).sum().item<float>() / y.size(0);
-    }
-    float BaseClassifier::score(vector<vector<int>>& X, vector<int>& y)
-    {
-        if (!fitted) {
-            throw logic_error("Classifier has not been fitted");
-        }
-        auto m_ = X[0].size();
-        auto n_ = X.size();
-        vector<vector<int>> Xd(n_, vector<int>(m_, 0));
-        for (auto i = 0; i < n_; i++) {
-            Xd[i] = vector<int>(X[i].begin(), X[i].end());
-        }
-        return model.score(Xd, y);
-    }
-    vector<string> BaseClassifier::show()
-    {
-        return model.show();
-    }
-    void BaseClassifier::addNodes()
-    {
-        // Add all nodes to the network
-        for (auto feature : features) {
-            model.addNode(feature, states[feature].size());
-        }
-        model.addNode(className, states[className].size());
-    }
-}

src/BaseClassifier.h

@@ -1,46 +0,0 @@
-#ifndef CLASSIFIERS_H
-#define CLASSIFIERS_H
-#include <torch/torch.h>
-#include "Network.h"
-#include "Metrics.hpp"
-using namespace std;
-using namespace torch;
-
-namespace bayesnet {
-    class BaseClassifier {
-    private:
-        bool fitted;
-        BaseClassifier& build(vector<string>& features, string className, map<string, vector<int>>& states);
-    protected:
-        Network model;
-        int m, n; // m: number of samples, n: number of features
-        Tensor X;
-        vector<vector<int>> Xv;
-        Tensor y;
-        vector<int> yv;
-        Tensor dataset;
-        Metrics metrics;
-        vector<string> features;
-        string className;
-        map<string, vector<int>> states;
-        void checkFitParameters();
-        virtual void train() = 0;
-    public:
-        BaseClassifier(Network model);
-        virtual ~BaseClassifier() = default;
-        BaseClassifier& fit(vector<vector<int>>& X, vector<int>& y, vector<string>& features, string className, map<string, vector<int>>& states);
-        void addNodes();
-        Tensor predict(Tensor& X);
-        vector<int> predict(vector<vector<int>>& X);
-        float score(Tensor& X, Tensor& y);
-        float score(vector<vector<int>>& X, vector<int>& y);
-        vector<string> show();
-        virtual vector<string> graph(string title) = 0;
-    };
-}
-#endif
-
-
-
-
-

src/BayesNet/AODE.cc

@@ -2,14 +2,16 @@
 
 namespace bayesnet {
     AODE::AODE() : Ensemble() {}
-    void AODE::train()
+    void AODE::buildModel(const torch::Tensor& weights)
     {
         models.clear();
         for (int i = 0; i < features.size(); ++i) {
             models.push_back(std::make_unique<SPODE>(i));
         }
+        n_models = models.size();
+        significanceModels = vector<double>(n_models, 1.0);
     }
-    vector<string> AODE::graph(string title)
+    vector<string> AODE::graph(const string& title) const
     {
         return Ensemble::graph(title);
     }

src/BayesNet/AODE.h

@@ -5,10 +5,11 @@
 namespace bayesnet {
     class AODE : public Ensemble {
     protected:
-        void train() override;
+        void buildModel(const torch::Tensor& weights) override;
     public:
         AODE();
-        vector<string> graph(string title = "AODE");
+        virtual ~AODE() {};
+        vector<string> graph(const string& title = "AODE") const override;
     };
 }
 #endif

src/BayesNet/AODELd.cc

@@ -0,0 +1,41 @@
+#include "AODELd.h"
+#include "Models.h"
+
+namespace bayesnet {
+    using namespace std;
+    AODELd::AODELd() : Ensemble(), Proposal(dataset, features, className) {}
+    AODELd& AODELd::fit(torch::Tensor& X_, torch::Tensor& y_, const vector<string>& features_, const string& className_, map<string, vector<int>>& states_)
+    {
+        checkInput(X_, y_);
+        features = features_;
+        className = className_;
+        Xf = X_;
+        y = y_;
+        // Fills vectors Xv & yv with the data from tensors X_ (discretized) & y
+        states = fit_local_discretization(y);
+        // We have discretized the input data
+        // 1st we need to fit the model to build the normal TAN structure, TAN::fit initializes the base Bayesian network
+        Ensemble::fit(dataset, features, className, states);
+        return *this;
+
+    }
+    void AODELd::buildModel(const torch::Tensor& weights)
+    {
+        models.clear();
+        for (int i = 0; i < features.size(); ++i) {
+            models.push_back(std::make_unique<SPODELd>(i));
+        }
+        n_models = models.size();
+        significanceModels = vector<double>(n_models, 1.0);
+    }
+    void AODELd::trainModel(const torch::Tensor& weights)
+    {
+        for (const auto& model : models) {
+            model->fit(Xf, y, features, className, states);
+        }
+    }
+    vector<string> AODELd::graph(const string& name) const
+    {
+        return Ensemble::graph(name);
+    }
+}

src/BayesNet/AODELd.h

@@ -0,0 +1,21 @@
+#ifndef AODELD_H
+#define AODELD_H
+#include "Ensemble.h"
+#include "Proposal.h"
+#include "SPODELd.h"
+
+namespace bayesnet {
+    using namespace std;
+    class AODELd : public Ensemble, public Proposal {
+    protected:
+        void trainModel(const torch::Tensor& weights) override;
+        void buildModel(const torch::Tensor& weights) override;
+    public:
+        AODELd();
+        AODELd& fit(torch::Tensor& X_, torch::Tensor& y_, const vector<string>& features_, const string& className_, map<string, vector<int>>& states_) override;
+        virtual ~AODELd() = default;
+        vector<string> graph(const string& name = "AODELd") const override;
+        static inline string version() { return "0.0.1"; };
+    };
+}
+#endif // !AODELD_H

src/BayesNet/BaseClassifier.h

@@ -0,0 +1,36 @@
+#ifndef BASE_H
+#define BASE_H
+#include <torch/torch.h>
+#include <nlohmann/json.hpp>
+#include <vector>
+namespace bayesnet {
+    using namespace std;
+    enum status_t { NORMAL, WARNING, ERROR };
+    class BaseClassifier {
+    protected:
+        virtual void trainModel(const torch::Tensor& weights) = 0;
+    public:
+        // X is nxm vector, y is nx1 vector
+        virtual BaseClassifier& fit(vector<vector<int>>& X, vector<int>& y, const vector<string>& features, const string& className, map<string, vector<int>>& states) = 0;
+        // X is nxm tensor, y is nx1 tensor
+        virtual BaseClassifier& fit(torch::Tensor& X, torch::Tensor& y, const vector<string>& features, const string& className, map<string, vector<int>>& states) = 0;
+        virtual BaseClassifier& fit(torch::Tensor& dataset, const vector<string>& features, const string& className, map<string, vector<int>>& states) = 0;
+        virtual BaseClassifier& fit(torch::Tensor& dataset, const vector<string>& features, const string& className, map<string, vector<int>>& states, const torch::Tensor& weights) = 0;
+        virtual ~BaseClassifier() = default;
+        torch::Tensor virtual predict(torch::Tensor& X) = 0;
+        vector<int> virtual predict(vector<vector<int>>& X) = 0;
+        status_t virtual getStatus() const = 0;
+        float virtual score(vector<vector<int>>& X, vector<int>& y) = 0;
+        float virtual score(torch::Tensor& X, torch::Tensor& y) = 0;
+        int virtual getNumberOfNodes()const = 0;
+        int virtual getNumberOfEdges()const = 0;
+        int virtual getNumberOfStates() const = 0;
+        vector<string> virtual show() const = 0;
+        vector<string> virtual graph(const string& title = "") const = 0;
+        const string inline getVersion() const { return "0.2.0"; };
+        vector<string> virtual topological_order() = 0;
+        void virtual dump_cpt()const = 0;
+        virtual void setHyperparameters(nlohmann::json& hyperparameters) = 0;
+    };
+}
+#endif

src/BayesNet/BayesMetrics.cc

@@ -0,0 +1,163 @@
+#include "BayesMetrics.h"
+#include "Mst.h"
+namespace bayesnet {
+    //samples is n+1xm tensor used to fit the model
+    Metrics::Metrics(const torch::Tensor& samples, const vector<string>& features, const string& className, const int classNumStates)
+        : samples(samples)
+        , features(features)
+        , className(className)
+        , classNumStates(classNumStates)
+    {
+    }
+    //samples is nxm vector used to fit the model
+    Metrics::Metrics(const vector<vector<int>>& vsamples, const vector<int>& labels, const vector<string>& features, const string& className, const int classNumStates)
+        : features(features)
+        , className(className)
+        , classNumStates(classNumStates)
+        , samples(torch::zeros({ static_cast<int>(vsamples[0].size()), static_cast<int>(vsamples.size() + 1) }, torch::kInt32))
+    {
+        for (int i = 0; i < vsamples.size(); ++i) {
+            samples.index_put_({ i,  "..." }, torch::tensor(vsamples[i], torch::kInt32));
+        }
+        samples.index_put_({ -1, "..." }, torch::tensor(labels, torch::kInt32));
+    }
+    vector<int> Metrics::SelectKBestWeighted(const torch::Tensor& weights, bool ascending, unsigned k)
+    {
+        // Return the K Best features 
+        auto n = samples.size(0) - 1;
+        if (k == 0) {
+            k = n;
+        }
+        // compute scores
+        scoresKBest.clear();
+        featuresKBest.clear();
+        auto label = samples.index({ -1, "..." });
+        for (int i = 0; i < n; ++i) {
+            scoresKBest.push_back(mutualInformation(label, samples.index({ i, "..." }), weights));
+            featuresKBest.push_back(i);
+        }
+        // sort & reduce scores and features
+        if (ascending) {
+            sort(featuresKBest.begin(), featuresKBest.end(), [&](int i, int j)
+                { return scoresKBest[i] < scoresKBest[j]; });
+            sort(scoresKBest.begin(), scoresKBest.end(), std::less<double>());
+            if (k < n) {
+                for (int i = 0; i < n - k; ++i) {
+                    featuresKBest.erase(featuresKBest.begin());
+                    scoresKBest.erase(scoresKBest.begin());
+                }
+            }
+        } else {
+            sort(featuresKBest.begin(), featuresKBest.end(), [&](int i, int j)
+                { return scoresKBest[i] > scoresKBest[j]; });
+            sort(scoresKBest.begin(), scoresKBest.end(), std::greater<double>());
+            featuresKBest.resize(k);
+            scoresKBest.resize(k);
+        }
+        return featuresKBest;
+    }
+    vector<double> Metrics::getScoresKBest() const
+    {
+        return scoresKBest;
+    }
+
+    torch::Tensor Metrics::conditionalEdge(const torch::Tensor& weights)
+    {
+        auto result = vector<double>();
+        auto source = vector<string>(features);
+        source.push_back(className);
+        auto combinations = doCombinations(source);
+        // Compute class prior
+        auto margin = torch::zeros({ classNumStates }, torch::kFloat);
+        for (int value = 0; value < classNumStates; ++value) {
+            auto mask = samples.index({ -1,  "..." }) == value;
+            margin[value] = mask.sum().item<double>() / samples.size(1);
+        }
+        for (auto [first, second] : combinations) {
+            int index_first = find(features.begin(), features.end(), first) - features.begin();
+            int index_second = find(features.begin(), features.end(), second) - features.begin();
+            double accumulated = 0;
+            for (int value = 0; value < classNumStates; ++value) {
+                auto mask = samples.index({ -1, "..." }) == value;
+                auto first_dataset = samples.index({ index_first, mask });
+                auto second_dataset = samples.index({ index_second, mask });
+                auto weights_dataset = weights.index({ mask });
+                auto mi = mutualInformation(first_dataset, second_dataset, weights_dataset);
+                auto pb = margin[value].item<double>();
+                accumulated += pb * mi;
+            }
+            result.push_back(accumulated);
+        }
+        long n_vars = source.size();
+        auto matrix = torch::zeros({ n_vars, n_vars });
+        auto indices = torch::triu_indices(n_vars, n_vars, 1);
+        for (auto i = 0; i < result.size(); ++i) {
+            auto x = indices[0][i];
+            auto y = indices[1][i];
+            matrix[x][y] = result[i];
+            matrix[y][x] = result[i];
+        }
+        return matrix;
+    }
+    // To use in Python
+    vector<float> Metrics::conditionalEdgeWeights(vector<float>& weights_)
+    {
+        const torch::Tensor weights = torch::tensor(weights_);
+        auto matrix = conditionalEdge(weights);
+        std::vector<float> v(matrix.data_ptr<float>(), matrix.data_ptr<float>() + matrix.numel());
+        return v;
+    }
+    double Metrics::entropy(const torch::Tensor& feature, const torch::Tensor& weights)
+    {
+        torch::Tensor counts = feature.bincount(weights);
+        double totalWeight = counts.sum().item<double>();
+        torch::Tensor probs = counts.to(torch::kFloat) / totalWeight;
+        torch::Tensor logProbs = torch::log(probs);
+        torch::Tensor entropy = -probs * logProbs;
+        return entropy.nansum().item<double>();
+    }
+    // H(Y|X) = sum_{x in X} p(x) H(Y|X=x)
+    double Metrics::conditionalEntropy(const torch::Tensor& firstFeature, const torch::Tensor& secondFeature, const torch::Tensor& weights)
+    {
+        int numSamples = firstFeature.sizes()[0];
+        torch::Tensor featureCounts = secondFeature.bincount(weights);
+        unordered_map<int, unordered_map<int, double>> jointCounts;
+        double totalWeight = 0;
+        for (auto i = 0; i < numSamples; i++) {
+            jointCounts[secondFeature[i].item<int>()][firstFeature[i].item<int>()] += weights[i].item<double>();
+            totalWeight += weights[i].item<float>();
+        }
+        if (totalWeight == 0)
+            return 0;
+        double entropyValue = 0;
+        for (int value = 0; value < featureCounts.sizes()[0]; ++value) {
+            double p_f = featureCounts[value].item<double>() / totalWeight;
+            double entropy_f = 0;
+            for (auto& [label, jointCount] : jointCounts[value]) {
+                double p_l_f = jointCount / featureCounts[value].item<double>();
+                if (p_l_f > 0) {
+                    entropy_f -= p_l_f * log(p_l_f);
+                } else {
+                    entropy_f = 0;
+                }
+            }
+            entropyValue += p_f * entropy_f;
+        }
+        return entropyValue;
+    }
+    // I(X;Y) = H(Y) - H(Y|X)
+    double Metrics::mutualInformation(const torch::Tensor& firstFeature, const torch::Tensor& secondFeature, const torch::Tensor& weights)
+    {
+        return entropy(firstFeature, weights) - conditionalEntropy(firstFeature, secondFeature, weights);
+    }
+    /*
+    Compute the maximum spanning tree considering the weights as distances
+    and the indices of the weights as nodes of this square matrix using
+    Kruskal algorithm
+    */
+    vector<pair<int, int>> Metrics::maximumSpanningTree(const vector<string>& features, const Tensor& weights, const int root)
+    {
+        auto mst = MST(features, weights, root);
+        return mst.maximumSpanningTree();
+    }
+}

src/BayesNet/BayesMetrics.h

@@ -0,0 +1,51 @@
+#ifndef BAYESNET_METRICS_H
+#define BAYESNET_METRICS_H
+#include <torch/torch.h>
+#include <vector>
+#include <string>
+namespace bayesnet {
+    using namespace std;
+    using namespace torch;
+    class Metrics {
+    private:
+        int classNumStates = 0;
+        vector<double> scoresKBest;
+        vector<int> featuresKBest; // sorted indices of the features
+        double conditionalEntropy(const Tensor& firstFeature, const Tensor& secondFeature, const Tensor& weights);
+    protected:
+        Tensor samples; // n+1xm tensor used to fit the model where samples[-1] is the y vector
+        string className;
+        double entropy(const Tensor& feature, const Tensor& weights);
+        vector<string> features;
+        template <class T>
+        vector<pair<T, T>> doCombinations(const vector<T>& source)
+        {
+            vector<pair<T, T>> result;
+            for (int i = 0; i < source.size(); ++i) {
+                T temp = source[i];
+                for (int j = i + 1; j < source.size(); ++j) {
+                    result.push_back({ temp, source[j] });
+                }
+            }
+            return result;
+        }
+        template <class T>
+        T pop_first(vector<T>& v)
+        {
+            T temp = v[0];
+            v.erase(v.begin());
+            return temp;
+        }
+    public:
+        Metrics() = default;
+        Metrics(const torch::Tensor& samples, const vector<string>& features, const string& className, const int classNumStates);
+        Metrics(const vector<vector<int>>& vsamples, const vector<int>& labels, const vector<string>& features, const string& className, const int classNumStates);
+        vector<int> SelectKBestWeighted(const torch::Tensor& weights, bool ascending = false, unsigned k = 0);
+        vector<double> getScoresKBest() const;
+        double mutualInformation(const Tensor& firstFeature, const Tensor& secondFeature, const Tensor& weights);
+        vector<float> conditionalEdgeWeights(vector<float>& weights); // To use in Python
+        Tensor conditionalEdge(const torch::Tensor& weights);
+        vector<pair<int, int>> maximumSpanningTree(const vector<string>& features, const Tensor& weights, const int root);
+    };
+}
+#endif

src/BayesNet/BoostAODE.cc

@@ -0,0 +1,195 @@
+#include <set>
+#include <functional>
+#include <limits.h>
+#include "BoostAODE.h"
+#include "Colors.h"
+#include "Folding.h"
+#include "Paths.h"
+#include "CFS.h"
+#include "FCBF.h"
+#include "IWSS.h"
+
+namespace bayesnet {
+    BoostAODE::BoostAODE() : Ensemble() {}
+    void BoostAODE::buildModel(const torch::Tensor& weights)
+    {
+        // Models shall be built in trainModel
+        models.clear();
+        n_models = 0;
+        // Prepare the validation dataset
+        auto y_ = dataset.index({ -1, "..." });
+        if (convergence) {
+            // Prepare train & validation sets from train data
+            auto fold = platform::StratifiedKFold(5, y_, 271);
+            dataset_ = torch::clone(dataset);
+            // save input dataset
+            auto [train, test] = fold.getFold(0);
+            auto train_t = torch::tensor(train);
+            auto test_t = torch::tensor(test);
+            // Get train and validation sets
+            X_train = dataset.index({ torch::indexing::Slice(0, dataset.size(0) - 1), train_t });
+            y_train = dataset.index({ -1, train_t });
+            X_test = dataset.index({ torch::indexing::Slice(0, dataset.size(0) - 1), test_t });
+            y_test = dataset.index({ -1, test_t });
+            dataset = X_train;
+            m = X_train.size(1);
+            auto n_classes = states.at(className).size();
+            metrics = Metrics(dataset, features, className, n_classes);
+            // Build dataset with train data
+            buildDataset(y_train);
+        } else {
+            // Use all data to train
+            X_train = dataset.index({ torch::indexing::Slice(0, dataset.size(0) - 1), "..." });
+            y_train = y_;
+        }
+    }
+    void BoostAODE::setHyperparameters(nlohmann::json& hyperparameters)
+    {
+        // Check if hyperparameters are valid
+        const vector<string> validKeys = { "repeatSparent", "maxModels", "ascending", "convergence", "threshold", "select_features" };
+        checkHyperparameters(validKeys, hyperparameters);
+        if (hyperparameters.contains("repeatSparent")) {
+            repeatSparent = hyperparameters["repeatSparent"];
+        }
+        if (hyperparameters.contains("maxModels")) {
+            maxModels = hyperparameters["maxModels"];
+        }
+        if (hyperparameters.contains("ascending")) {
+            ascending = hyperparameters["ascending"];
+        }
+        if (hyperparameters.contains("convergence")) {
+            convergence = hyperparameters["convergence"];
+        }
+        if (hyperparameters.contains("threshold")) {
+            threshold = hyperparameters["threshold"];
+        }
+        if (hyperparameters.contains("select_features")) {
+            auto selectedAlgorithm = hyperparameters["select_features"];
+            vector<string> algos = { "IWSS", "FCBF", "CFS" };
+            selectFeatures = true;
+            algorithm = selectedAlgorithm;
+            if (find(algos.begin(), algos.end(), selectedAlgorithm) == algos.end()) {
+                throw invalid_argument("Invalid selectFeatures value [IWSS, FCBF, CFS]");
+            }
+        }
+    }
+    unordered_set<int> BoostAODE::initializeModels()
+    {
+        unordered_set<int> featuresUsed;
+        Tensor weights_ = torch::full({ m }, 1.0 / m, torch::kFloat64);
+        int maxFeatures = 0;
+        if (algorithm == "CFS") {
+            featureSelector = new CFS(dataset, features, className, maxFeatures, states.at(className).size(), weights_);
+        } else if (algorithm == "IWSS") {
+            if (threshold < 0 || threshold >0.5) {
+                throw invalid_argument("Invalid threshold value for IWSS [0, 0.5]");
+            }
+            featureSelector = new IWSS(dataset, features, className, maxFeatures, states.at(className).size(), weights_, threshold);
+        } else if (algorithm == "FCBF") {
+            if (threshold < 1e-7 || threshold > 1) {
+                throw invalid_argument("Invalid threshold value [1e-7, 1]");
+            }
+            featureSelector = new FCBF(dataset, features, className, maxFeatures, states.at(className).size(), weights_, threshold);
+        }
+        featureSelector->fit();
+        auto cfsFeatures = featureSelector->getFeatures();
+        for (const int& feature : cfsFeatures) {
+            // cout << "Feature: [" << feature << "] " << feature << " " << features.at(feature) << endl;
+            featuresUsed.insert(feature);
+            unique_ptr<Classifier> model = std::make_unique<SPODE>(feature);
+            model->fit(dataset, features, className, states, weights_);
+            models.push_back(std::move(model));
+            significanceModels.push_back(1.0);
+            n_models++;
+        }
+        delete featureSelector;
+        return featuresUsed;
+    }
+    void BoostAODE::trainModel(const torch::Tensor& weights)
+    {
+        unordered_set<int> featuresUsed;
+        if (selectFeatures) {
+            featuresUsed = initializeModels();
+        }
+        if (maxModels == 0)
+            maxModels = .1 * n > 10 ? .1 * n : n;
+        Tensor weights_ = torch::full({ m }, 1.0 / m, torch::kFloat64);
+        bool exitCondition = false;
+        // Variables to control the accuracy finish condition
+        double priorAccuracy = 0.0;
+        double delta = 1.0;
+        double threshold = 1e-4;
+        int tolerance = 5; // number of times the accuracy can be lower than the threshold
+        int count = 0; // number of times the accuracy is lower than the threshold
+        fitted = true; // to enable predict
+        // Step 0: Set the finish condition
+        // if not repeatSparent a finish condition is run out of features
+        // n_models == maxModels
+        // epsiolon sub t > 0.5 => inverse the weights policy
+        // validation error is not decreasing
+        while (!exitCondition) {
+            // Step 1: Build ranking with mutual information
+            auto featureSelection = metrics.SelectKBestWeighted(weights_, ascending, n); // Get all the features sorted
+            unique_ptr<Classifier> model;
+            auto feature = featureSelection[0];
+            if (!repeatSparent || featuresUsed.size() < featureSelection.size()) {
+                bool used = true;
+                for (const auto& feat : featureSelection) {
+                    if (find(featuresUsed.begin(), featuresUsed.end(), feat) != featuresUsed.end()) {
+                        continue;
+                    }
+                    used = false;
+                    feature = feat;
+                    break;
+                }
+                if (used) {
+                    exitCondition = true;
+                    continue;
+                }
+            }
+            featuresUsed.insert(feature);
+            model = std::make_unique<SPODE>(feature);
+            model->fit(dataset, features, className, states, weights_);
+            auto ypred = model->predict(X_train);
+            // Step 3.1: Compute the classifier amout of say
+            auto mask_wrong = ypred != y_train;
+            auto mask_right = ypred == y_train;
+            auto masked_weights = weights_ * mask_wrong.to(weights_.dtype());
+            double epsilon_t = masked_weights.sum().item<double>();
+            double wt = (1 - epsilon_t) / epsilon_t;
+            double alpha_t = epsilon_t == 0 ? 1 : 0.5 * log(wt);
+            // Step 3.2: Update weights for next classifier
+            // Step 3.2.1: Update weights of wrong samples
+            weights_ += mask_wrong.to(weights_.dtype()) * exp(alpha_t) * weights_;
+            // Step 3.2.2: Update weights of right samples
+            weights_ += mask_right.to(weights_.dtype()) * exp(-alpha_t) * weights_;
+            // Step 3.3: Normalise the weights
+            double totalWeights = torch::sum(weights_).item<double>();
+            weights_ = weights_ / totalWeights;
+            // Step 3.4: Store classifier and its accuracy to weigh its future vote
+            models.push_back(std::move(model));
+            significanceModels.push_back(alpha_t);
+            n_models++;
+            if (convergence) {
+                auto y_val_predict = predict(X_test);
+                double accuracy = (y_val_predict == y_test).sum().item<double>() / (double)y_test.size(0);
+                if (priorAccuracy == 0) {
+                    priorAccuracy = accuracy;
+                } else {
+                    delta = accuracy - priorAccuracy;
+                }
+                if (delta < threshold) {
+                    count++;
+                }
+            }
+            exitCondition = n_models >= maxModels && repeatSparent || epsilon_t > 0.5 || count > tolerance;
+        }
+        if (featuresUsed.size() != features.size()) {
+            status = WARNING;
+        }
+    }
+    vector<string> BoostAODE::graph(const string& title) const
+    {
+        return Ensemble::graph(title);
+    }
+}

src/BayesNet/BoostAODE.h

@@ -0,0 +1,32 @@
+#ifndef BOOSTAODE_H
+#define BOOSTAODE_H
+#include "Ensemble.h"
+#include <map>
+#include "SPODE.h"
+#include "FeatureSelect.h"
+namespace bayesnet {
+    class BoostAODE : public Ensemble {
+    public:
+        BoostAODE();
+        virtual ~BoostAODE() {};
+        vector<string> graph(const string& title = "BoostAODE") const override;
+        void setHyperparameters(nlohmann::json& hyperparameters) override;
+    protected:
+        void buildModel(const torch::Tensor& weights) override;
+        void trainModel(const torch::Tensor& weights) override;
+    private:
+        torch::Tensor dataset_;
+        torch::Tensor X_train, y_train, X_test, y_test;
+        unordered_set<int> initializeModels();
+        // Hyperparameters
+        bool repeatSparent = false; // if true, a feature can be selected more than once
+        int maxModels = 0;
+        bool ascending = false; //Process KBest features ascending or descending order
+        bool convergence = false; //if true, stop when the model does not improve
+        bool selectFeatures = false; // if true, use feature selection
+        string algorithm = ""; // Selected feature selection algorithm
+        FeatureSelect* featureSelector = nullptr;
+        double threshold = -1;
+    };
+}
+#endif

src/BayesNet/CFS.cc

@@ -0,0 +1,72 @@
+#include "CFS.h"
+#include <limits>
+#include "bayesnetUtils.h"
+namespace bayesnet {
+    void CFS::fit()
+    {
+        initialize();
+        computeSuLabels();
+        auto featureOrder = argsort(suLabels); // sort descending order
+        auto continueCondition = true;
+        auto feature = featureOrder[0];
+        selectedFeatures.push_back(feature);
+        selectedScores.push_back(suLabels[feature]);
+        selectedFeatures.erase(selectedFeatures.begin());
+        while (continueCondition) {
+            double merit = numeric_limits<double>::lowest();
+            int bestFeature = -1;
+            for (auto feature : featureOrder) {
+                selectedFeatures.push_back(feature);
+                // Compute merit with selectedFeatures
+                auto meritNew = computeMeritCFS();
+                if (meritNew > merit) {
+                    merit = meritNew;
+                    bestFeature = feature;
+                }
+                selectedFeatures.pop_back();
+            }
+            if (bestFeature == -1) {
+                // meritNew has to be nan due to constant features
+                break;
+            }
+            selectedFeatures.push_back(bestFeature);
+            selectedScores.push_back(merit);
+            featureOrder.erase(remove(featureOrder.begin(), featureOrder.end(), bestFeature), featureOrder.end());
+            continueCondition = computeContinueCondition(featureOrder);
+        }
+        fitted = true;
+    }
+    bool CFS::computeContinueCondition(const vector<int>& featureOrder)
+    {
+        if (selectedFeatures.size() == maxFeatures || featureOrder.size() == 0) {
+            return false;
+        }
+        if (selectedScores.size() >= 5) {
+            /*
+            "To prevent the best first search from exploring the entire
+            feature subset search space, a stopping criterion is imposed.
+            The search will terminate if five consecutive fully expanded
+            subsets show no improvement over the current best subset."
+            as stated in Mark A.Hall Thesis
+            */
+            double item_ant = numeric_limits<double>::lowest();
+            int num = 0;
+            vector<double> lastFive(selectedScores.end() - 5, selectedScores.end());
+            for (auto item : lastFive) {
+                if (item_ant == numeric_limits<double>::lowest()) {
+                    item_ant = item;
+                }
+                if (item > item_ant) {
+                    break;
+                } else {
+                    num++;
+                    item_ant = item;
+                }
+            }
+            if (num == 5) {
+                return false;
+            }
+        }
+        return true;
+    }
+}

src/BayesNet/CFS.h

@@ -0,0 +1,21 @@
+#ifndef CFS_H
+#define CFS_H
+#include <torch/torch.h>
+#include <vector>
+#include "FeatureSelect.h"
+using namespace std;
+namespace bayesnet {
+    class CFS : public FeatureSelect {
+    public:
+        // dataset is a n+1xm tensor of integers where dataset[-1] is the y vector
+        CFS(const torch::Tensor& samples, const vector<string>& features, const string& className, const int maxFeatures, const int classNumStates, const torch::Tensor& weights) :
+            FeatureSelect(samples, features, className, maxFeatures, classNumStates, weights)
+        {
+        }
+        virtual ~CFS() {};
+        void fit() override;
+    private:
+        bool computeContinueCondition(const vector<int>& featureOrder);
+    };
+}
+#endif

src/BayesNet/CMakeLists.txt

@@ -0,0 +1,9 @@
+include_directories(${BayesNet_SOURCE_DIR}/lib/mdlp)
+include_directories(${BayesNet_SOURCE_DIR}/lib/Files)
+include_directories(${BayesNet_SOURCE_DIR}/lib/json/include)
+include_directories(${BayesNet_SOURCE_DIR}/src/BayesNet)
+include_directories(${BayesNet_SOURCE_DIR}/src/Platform)
+add_library(BayesNet bayesnetUtils.cc Network.cc Node.cc BayesMetrics.cc Classifier.cc 
+    KDB.cc TAN.cc SPODE.cc Ensemble.cc AODE.cc TANLd.cc KDBLd.cc SPODELd.cc AODELd.cc BoostAODE.cc 
+    Mst.cc Proposal.cc CFS.cc FCBF.cc IWSS.cc FeatureSelect.cc ${BayesNet_SOURCE_DIR}/src/Platform/Models.cc)
+target_link_libraries(BayesNet mdlp "${TORCH_LIBRARIES}")

src/BayesNet/Classifier.cc

@@ -0,0 +1,172 @@
+#include "Classifier.h"
+#include "bayesnetUtils.h"
+
+namespace bayesnet {
+    using namespace torch;
+
+    Classifier::Classifier(Network model) : model(model), m(0), n(0), metrics(Metrics()), fitted(false) {}
+    Classifier& Classifier::build(const vector<string>& features, const string& className, map<string, vector<int>>& states, const torch::Tensor& weights)
+    {
+        this->features = features;
+        this->className = className;
+        this->states = states;
+        m = dataset.size(1);
+        n = dataset.size(0) - 1;
+        checkFitParameters();
+        auto n_classes = states.at(className).size();
+        metrics = Metrics(dataset, features, className, n_classes);
+        model.initialize();
+        buildModel(weights);
+        trainModel(weights);
+        fitted = true;
+        return *this;
+    }
+    void Classifier::buildDataset(Tensor& ytmp)
+    {
+        try {
+            auto yresized = torch::transpose(ytmp.view({ ytmp.size(0), 1 }), 0, 1);
+            dataset = torch::cat({ dataset, yresized }, 0);
+        }
+        catch (const std::exception& e) {
+            std::cerr << e.what() << '\n';
+            cout << "X dimensions: " << dataset.sizes() << "\n";
+            cout << "y dimensions: " << ytmp.sizes() << "\n";
+            exit(1);
+        }
+    }
+    void Classifier::trainModel(const torch::Tensor& weights)
+    {
+        model.fit(dataset, weights, features, className, states);
+    }
+    // X is nxm where n is the number of features and m the number of samples
+    Classifier& Classifier::fit(torch::Tensor& X, torch::Tensor& y, const vector<string>& features, const string& className, map<string, vector<int>>& states)
+    {
+        dataset = X;
+        buildDataset(y);
+        const torch::Tensor weights = torch::full({ dataset.size(1) }, 1.0 / dataset.size(1), torch::kDouble);
+        return build(features, className, states, weights);
+    }
+    // X is nxm where n is the number of features and m the number of samples
+    Classifier& Classifier::fit(vector<vector<int>>& X, vector<int>& y, const vector<string>& features, const string& className, map<string, vector<int>>& states)
+    {
+        dataset = torch::zeros({ static_cast<int>(X.size()), static_cast<int>(X[0].size()) }, kInt32);
+        for (int i = 0; i < X.size(); ++i) {
+            dataset.index_put_({ i, "..." }, torch::tensor(X[i], kInt32));
+        }
+        auto ytmp = torch::tensor(y, kInt32);
+        buildDataset(ytmp);
+        const torch::Tensor weights = torch::full({ dataset.size(1) }, 1.0 / dataset.size(1), torch::kDouble);
+        return build(features, className, states, weights);
+    }
+    Classifier& Classifier::fit(torch::Tensor& dataset, const vector<string>& features, const string& className, map<string, vector<int>>& states)
+    {
+        this->dataset = dataset;
+        const torch::Tensor weights = torch::full({ dataset.size(1) }, 1.0 / dataset.size(1), torch::kDouble);
+        return build(features, className, states, weights);
+    }
+    Classifier& Classifier::fit(torch::Tensor& dataset, const vector<string>& features, const string& className, map<string, vector<int>>& states, const torch::Tensor& weights)
+    {
+        this->dataset = dataset;
+        return build(features, className, states, weights);
+    }
+    void Classifier::checkFitParameters()
+    {
+        if (torch::is_floating_point(dataset)) {
+            throw invalid_argument("dataset (X, y) must be of type Integer");
+        }
+        if (n != features.size()) {
+            throw invalid_argument("Classifier: X " + to_string(n) + " and features " + to_string(features.size()) + " must have the same number of features");
+        }
+        if (states.find(className) == states.end()) {
+            throw invalid_argument("className not found in states");
+        }
+        for (auto feature : features) {
+            if (states.find(feature) == states.end()) {
+                throw invalid_argument("feature [" + feature + "] not found in states");
+            }
+        }
+    }
+    Tensor Classifier::predict(Tensor& X)
+    {
+        if (!fitted) {
+            throw logic_error("Classifier has not been fitted");
+        }
+        return model.predict(X);
+    }
+    vector<int> Classifier::predict(vector<vector<int>>& X)
+    {
+        if (!fitted) {
+            throw logic_error("Classifier has not been fitted");
+        }
+        auto m_ = X[0].size();
+        auto n_ = X.size();
+        vector<vector<int>> Xd(n_, vector<int>(m_, 0));
+        for (auto i = 0; i < n_; i++) {
+            Xd[i] = vector<int>(X[i].begin(), X[i].end());
+        }
+        auto yp = model.predict(Xd);
+        return yp;
+    }
+    float Classifier::score(Tensor& X, Tensor& y)
+    {
+        if (!fitted) {
+            throw logic_error("Classifier has not been fitted");
+        }
+        Tensor y_pred = predict(X);
+        return (y_pred == y).sum().item<float>() / y.size(0);
+    }
+    float Classifier::score(vector<vector<int>>& X, vector<int>& y)
+    {
+        if (!fitted) {
+            throw logic_error("Classifier has not been fitted");
+        }
+        return model.score(X, y);
+    }
+    vector<string> Classifier::show() const
+    {
+        return model.show();
+    }
+    void Classifier::addNodes()
+    {
+        // Add all nodes to the network
+        for (const auto& feature : features) {
+            model.addNode(feature);
+        }
+        model.addNode(className);
+    }
+    int Classifier::getNumberOfNodes() const
+    {
+        // Features does not include class
+        return fitted ? model.getFeatures().size() + 1 : 0;
+    }
+    int Classifier::getNumberOfEdges() const
+    {
+        return fitted ? model.getNumEdges() : 0;
+    }
+    int Classifier::getNumberOfStates() const
+    {
+        return fitted ? model.getStates() : 0;
+    }
+    vector<string> Classifier::topological_order()
+    {
+        return model.topological_sort();
+    }
+    void Classifier::dump_cpt() const
+    {
+        model.dump_cpt();
+    }
+    void Classifier::checkHyperparameters(const vector<string>& validKeys, nlohmann::json& hyperparameters)
+    {
+        for (const auto& item : hyperparameters.items()) {
+            if (find(validKeys.begin(), validKeys.end(), item.key()) == validKeys.end()) {
+                throw invalid_argument("Hyperparameter " + item.key() + " is not valid");
+            }
+        }
+    }
+    void Classifier::setHyperparameters(nlohmann::json& hyperparameters)
+    {
+        // Check if hyperparameters are valid, default is no hyperparameters
+        const vector<string> validKeys = { };
+        checkHyperparameters(validKeys, hyperparameters);
+    }
+}

src/BayesNet/Classifier.h

@@ -0,0 +1,56 @@
+#ifndef CLASSIFIER_H
+#define CLASSIFIER_H
+#include <torch/torch.h>
+#include "BaseClassifier.h"
+#include "Network.h"
+#include "BayesMetrics.h"
+using namespace std;
+using namespace torch;
+
+namespace bayesnet {
+    class Classifier : public BaseClassifier {
+    private:
+        Classifier& build(const vector<string>& features, const string& className, map<string, vector<int>>& states, const torch::Tensor& weights);
+    protected:
+        bool fitted;
+        int m, n; // m: number of samples, n: number of features
+        Network model;
+        Metrics metrics;
+        vector<string> features;
+        string className;
+        map<string, vector<int>> states;
+        Tensor dataset; // (n+1)xm tensor
+        status_t status = NORMAL;
+        void checkFitParameters();
+        virtual void buildModel(const torch::Tensor& weights) = 0;
+        void trainModel(const torch::Tensor& weights) override;
+        void checkHyperparameters(const vector<string>& validKeys, nlohmann::json& hyperparameters);
+        void buildDataset(torch::Tensor& y);
+    public:
+        Classifier(Network model);
+        virtual ~Classifier() = default;
+        Classifier& fit(vector<vector<int>>& X, vector<int>& y, const vector<string>& features, const string& className, map<string, vector<int>>& states) override;
+        Classifier& fit(torch::Tensor& X, torch::Tensor& y, const vector<string>& features, const string& className, map<string, vector<int>>& states) override;
+        Classifier& fit(torch::Tensor& dataset, const vector<string>& features, const string& className, map<string, vector<int>>& states) override;
+        Classifier& fit(torch::Tensor& dataset, const vector<string>& features, const string& className, map<string, vector<int>>& states, const torch::Tensor& weights) override;
+        void addNodes();
+        int getNumberOfNodes() const override;
+        int getNumberOfEdges() const override;
+        int getNumberOfStates() const override;
+        Tensor predict(Tensor& X) override;
+        status_t getStatus() const override { return status; }
+        vector<int> predict(vector<vector<int>>& X) override;
+        float score(Tensor& X, Tensor& y) override;
+        float score(vector<vector<int>>& X, vector<int>& y) override;
+        vector<string> show() const override;
+        vector<string> topological_order()  override;
+        void dump_cpt() const override;
+        void setHyperparameters(nlohmann::json& hyperparameters) override;
+    };
+}
+#endif
+
+
+
+
+

src/BayesNet/Ensemble.cc

@@ -0,0 +1,142 @@
+#include "Ensemble.h"
+
+namespace bayesnet {
+    using namespace torch;
+
+    Ensemble::Ensemble() : Classifier(Network()), n_models(0) {}
+
+    void Ensemble::trainModel(const torch::Tensor& weights)
+    {
+        n_models = models.size();
+        for (auto i = 0; i < n_models; ++i) {
+            // fit with vectors
+            models[i]->fit(dataset, features, className, states);
+        }
+    }
+    vector<int> Ensemble::voting(Tensor& y_pred)
+    {
+        auto y_pred_ = y_pred.accessor<int, 2>();
+        vector<int> y_pred_final;
+        int numClasses = states.at(className).size();
+        // y_pred is m x n_models with the prediction of every model for each sample
+        for (int i = 0; i < y_pred.size(0); ++i) {
+            // votes store in each index (value of class) the significance added by each model
+            // i.e. votes[0] contains how much value has the value 0 of class. That value is generated by the models predictions
+            vector<double> votes(numClasses, 0.0);
+            for (int j = 0; j < n_models; ++j) {
+                votes[y_pred_[i][j]] += significanceModels.at(j);
+            }
+            // argsort in descending order
+            auto indices = argsort(votes);
+            y_pred_final.push_back(indices[0]);
+        }
+        return y_pred_final;
+    }
+    Tensor Ensemble::predict(Tensor& X)
+    {
+        if (!fitted) {
+            throw logic_error("Ensemble has not been fitted");
+        }
+        Tensor y_pred = torch::zeros({ X.size(1), n_models }, kInt32);
+        auto threads{ vector<thread>() };
+        mutex mtx;
+        for (auto i = 0; i < n_models; ++i) {
+            threads.push_back(thread([&, i]() {
+                auto ypredict = models[i]->predict(X);
+                lock_guard<mutex> lock(mtx);
+                y_pred.index_put_({ "...", i }, ypredict);
+                }));
+        }
+        for (auto& thread : threads) {
+            thread.join();
+        }
+        return torch::tensor(voting(y_pred));
+    }
+    vector<int> Ensemble::predict(vector<vector<int>>& X)
+    {
+        if (!fitted) {
+            throw logic_error("Ensemble has not been fitted");
+        }
+        long m_ = X[0].size();
+        long n_ = X.size();
+        vector<vector<int>> Xd(n_, vector<int>(m_, 0));
+        for (auto i = 0; i < n_; i++) {
+            Xd[i] = vector<int>(X[i].begin(), X[i].end());
+        }
+        Tensor y_pred = torch::zeros({ m_, n_models }, kInt32);
+        for (auto i = 0; i < n_models; ++i) {
+            y_pred.index_put_({ "...", i }, torch::tensor(models[i]->predict(Xd), kInt32));
+        }
+        return voting(y_pred);
+    }
+    float Ensemble::score(Tensor& X, Tensor& y)
+    {
+        if (!fitted) {
+            throw logic_error("Ensemble has not been fitted");
+        }
+        auto y_pred = predict(X);
+        int correct = 0;
+        for (int i = 0; i < y_pred.size(0); ++i) {
+            if (y_pred[i].item<int>() == y[i].item<int>()) {
+                correct++;
+            }
+        }
+        return (double)correct / y_pred.size(0);
+    }
+    float Ensemble::score(vector<vector<int>>& X, vector<int>& y)
+    {
+        if (!fitted) {
+            throw logic_error("Ensemble has not been fitted");
+        }
+        auto y_pred = predict(X);
+        int correct = 0;
+        for (int i = 0; i < y_pred.size(); ++i) {
+            if (y_pred[i] == y[i]) {
+                correct++;
+            }
+        }
+        return (double)correct / y_pred.size();
+    }
+    vector<string> Ensemble::show() const
+    {
+        auto result = vector<string>();
+        for (auto i = 0; i < n_models; ++i) {
+            auto res = models[i]->show();
+            result.insert(result.end(), res.begin(), res.end());
+        }
+        return result;
+    }
+    vector<string> Ensemble::graph(const string& title) const
+    {
+        auto result = vector<string>();
+        for (auto i = 0; i < n_models; ++i) {
+            auto res = models[i]->graph(title + "_" + to_string(i));
+            result.insert(result.end(), res.begin(), res.end());
+        }
+        return result;
+    }
+    int Ensemble::getNumberOfNodes() const
+    {
+        int nodes = 0;
+        for (auto i = 0; i < n_models; ++i) {
+            nodes += models[i]->getNumberOfNodes();
+        }
+        return nodes;
+    }
+    int Ensemble::getNumberOfEdges() const
+    {
+        int edges = 0;
+        for (auto i = 0; i < n_models; ++i) {
+            edges += models[i]->getNumberOfEdges();
+        }
+        return edges;
+    }
+    int Ensemble::getNumberOfStates() const
+    {
+        int nstates = 0;
+        for (auto i = 0; i < n_models; ++i) {
+            nstates += models[i]->getNumberOfStates();
+        }
+        return nstates;
+    }
+}

src/BayesNet/Ensemble.h

@@ -0,0 +1,41 @@
+#ifndef ENSEMBLE_H
+#define ENSEMBLE_H
+#include <torch/torch.h>
+#include "Classifier.h"
+#include "BayesMetrics.h"
+#include "bayesnetUtils.h"
+using namespace std;
+using namespace torch;
+
+namespace bayesnet {
+    class Ensemble : public Classifier {
+    private:
+        Ensemble& build(vector<string>& features, string className, map<string, vector<int>>& states);
+    protected:
+        unsigned n_models;
+        vector<unique_ptr<Classifier>> models;
+        vector<double> significanceModels;
+        void trainModel(const torch::Tensor& weights) override;
+        vector<int> voting(Tensor& y_pred);
+    public:
+        Ensemble();
+        virtual ~Ensemble() = default;
+        Tensor predict(Tensor& X) override;
+        vector<int> predict(vector<vector<int>>& X) override;
+        float score(Tensor& X, Tensor& y) override;
+        float score(vector<vector<int>>& X, vector<int>& y) override;
+        int getNumberOfNodes() const override;
+        int getNumberOfEdges() const override;
+        int getNumberOfStates() const override;
+        vector<string> show() const override;
+        vector<string> graph(const string& title) const override;
+        vector<string> topological_order()  override
+        {
+            return vector<string>();
+        }
+        void dump_cpt() const override
+        {
+        }
+    };
+}
+#endif

src/BayesNet/FCBF.cc

@@ -0,0 +1,44 @@
+#include "bayesnetUtils.h"
+#include "FCBF.h"
+namespace bayesnet {
+
+    FCBF::FCBF(const torch::Tensor& samples, const vector<string>& features, const string& className, const int maxFeatures, const int classNumStates, const torch::Tensor& weights, const double threshold) :
+        FeatureSelect(samples, features, className, maxFeatures, classNumStates, weights), threshold(threshold)
+    {
+        if (threshold < 1e-7) {
+            throw std::invalid_argument("Threshold cannot be less than 1e-7");
+        }
+    }
+    void FCBF::fit()
+    {
+        initialize();
+        computeSuLabels();
+        auto featureOrder = argsort(suLabels); // sort descending order
+        auto featureOrderCopy = featureOrder;
+        for (const auto& feature : featureOrder) {
+            // Don't self compare
+            featureOrderCopy.erase(featureOrderCopy.begin());
+            if (suLabels.at(feature) == 0.0) {
+                // The feature has been removed from the list
+                continue;
+            }
+            if (suLabels.at(feature) < threshold) {
+                break;
+            }
+            // Remove redundant features
+            for (const auto& featureCopy : featureOrderCopy) {
+                double value = computeSuFeatures(feature, featureCopy);
+                if (value >= suLabels.at(featureCopy)) {
+                    // Remove feature from list
+                    suLabels[featureCopy] = 0.0;
+                }
+            }
+            selectedFeatures.push_back(feature);
+            selectedScores.push_back(suLabels[feature]);
+            if (selectedFeatures.size() == maxFeatures) {
+                break;
+            }
+        }
+        fitted = true;
+    }
+}

src/BayesNet/FCBF.h

@@ -0,0 +1,18 @@
+#ifndef FCBF_H
+#define FCBF_H
+#include <torch/torch.h>
+#include <vector>
+#include "FeatureSelect.h"
+using namespace std;
+namespace bayesnet {
+    class FCBF : public FeatureSelect {
+    public:
+        // dataset is a n+1xm tensor of integers where dataset[-1] is the y vector
+        FCBF(const torch::Tensor& samples, const vector<string>& features, const string& className, const int maxFeatures, const int classNumStates, const torch::Tensor& weights, const double threshold);
+        virtual ~FCBF() {};
+        void fit() override;
+    private:
+        double threshold = -1;
+    };
+}
+#endif

src/BayesNet/FeatureSelect.cc

@@ -0,0 +1,79 @@
+#include "FeatureSelect.h"
+#include <limits>
+#include "bayesnetUtils.h"
+namespace bayesnet {
+    FeatureSelect::FeatureSelect(const torch::Tensor& samples, const vector<string>& features, const string& className, const int maxFeatures, const int classNumStates, const torch::Tensor& weights) :
+        Metrics(samples, features, className, classNumStates), maxFeatures(maxFeatures == 0 ? samples.size(0) - 1 : maxFeatures), weights(weights)
+
+    {
+    }
+    void FeatureSelect::initialize()
+    {
+        selectedFeatures.clear();
+        selectedScores.clear();
+    }
+    double FeatureSelect::symmetricalUncertainty(int a, int b)
+    {
+        /*
+        Compute symmetrical uncertainty. Normalize* information gain (mutual
+        information) with the entropies of the features in order to compensate
+        the bias due to high cardinality features. *Range [0, 1]
+        (https://www.sciencedirect.com/science/article/pii/S0020025519303603)
+        */
+        auto x = samples.index({ a, "..." });
+        auto y = samples.index({ b, "..." });
+        auto mu = mutualInformation(x, y, weights);
+        auto hx = entropy(x, weights);
+        auto hy = entropy(y, weights);
+        return 2.0 * mu / (hx + hy);
+    }
+    void FeatureSelect::computeSuLabels()
+    {
+        // Compute Simmetrical Uncertainty between features and labels
+        // https://en.wikipedia.org/wiki/Symmetric_uncertainty
+        for (int i = 0; i < features.size(); ++i) {
+            suLabels.push_back(symmetricalUncertainty(i, -1));
+        }
+    }
+    double FeatureSelect::computeSuFeatures(const int firstFeature, const int secondFeature)
+    {
+        // Compute Simmetrical Uncertainty between features
+        // https://en.wikipedia.org/wiki/Symmetric_uncertainty
+        try {
+            return suFeatures.at({ firstFeature, secondFeature });
+        }
+        catch (const out_of_range& e) {
+            double result = symmetricalUncertainty(firstFeature, secondFeature);
+            suFeatures[{firstFeature, secondFeature}] = result;
+            return result;
+        }
+    }
+    double FeatureSelect::computeMeritCFS()
+    {
+        double result;
+        double rcf = 0;
+        for (auto feature : selectedFeatures) {
+            rcf += suLabels[feature];
+        }
+        double rff = 0;
+        int n = selectedFeatures.size();
+        for (const auto& item : doCombinations(selectedFeatures)) {
+            rff += computeSuFeatures(item.first, item.second);
+        }
+        return rcf / sqrt(n + (n * n - n) * rff);
+    }
+    vector<int> FeatureSelect::getFeatures() const
+    {
+        if (!fitted) {
+            throw runtime_error("FeatureSelect not fitted");
+        }
+        return selectedFeatures;
+    }
+    vector<double> FeatureSelect::getScores() const
+    {
+        if (!fitted) {
+            throw runtime_error("FeatureSelect not fitted");
+        }
+        return selectedScores;
+    }
+}

src/BayesNet/FeatureSelect.h

@@ -0,0 +1,31 @@
+#ifndef FEATURE_SELECT_H
+#define FEATURE_SELECT_H
+#include <torch/torch.h>
+#include <vector>
+#include "BayesMetrics.h"
+using namespace std;
+namespace bayesnet {
+    class FeatureSelect : public Metrics {
+    public:
+        // dataset is a n+1xm tensor of integers where dataset[-1] is the y vector
+        FeatureSelect(const torch::Tensor& samples, const vector<string>& features, const string& className, const int maxFeatures, const int classNumStates, const torch::Tensor& weights);
+        virtual ~FeatureSelect() {};
+        virtual void fit() = 0;
+        vector<int> getFeatures() const;
+        vector<double> getScores() const;
+    protected:
+        void initialize();
+        void computeSuLabels();
+        double computeSuFeatures(const int a, const int b);
+        double symmetricalUncertainty(int a, int b);
+        double computeMeritCFS();
+        const torch::Tensor& weights;
+        int maxFeatures;
+        vector<int> selectedFeatures;
+        vector<double> selectedScores;
+        vector<double> suLabels;
+        map<pair<int, int>, double> suFeatures;
+        bool fitted = false;
+    };
+}
+#endif

src/BayesNet/IWSS.cc

@@ -0,0 +1,47 @@
+#include "IWSS.h"
+#include <limits>
+#include "bayesnetUtils.h"
+namespace bayesnet {
+    IWSS::IWSS(const torch::Tensor& samples, const vector<string>& features, const string& className, const int maxFeatures, const int classNumStates, const torch::Tensor& weights, const double threshold) :
+        FeatureSelect(samples, features, className, maxFeatures, classNumStates, weights), threshold(threshold)
+    {
+        if (threshold < 0 || threshold > .5) {
+            throw std::invalid_argument("Threshold has to be in [0, 0.5]");
+        }
+    }
+    void IWSS::fit()
+    {
+        initialize();
+        computeSuLabels();
+        auto featureOrder = argsort(suLabels); // sort descending order
+        auto featureOrderCopy = featureOrder;
+        // Add first and second features to result
+        //     First with its own score
+        auto first_feature = pop_first(featureOrderCopy);
+        selectedFeatures.push_back(first_feature);
+        selectedScores.push_back(suLabels.at(first_feature));
+        //     Second with the score of the candidates
+        selectedFeatures.push_back(pop_first(featureOrderCopy));
+        auto merit = computeMeritCFS();
+        selectedScores.push_back(merit);
+        for (const auto feature : featureOrderCopy) {
+            selectedFeatures.push_back(feature);
+            // Compute merit with selectedFeatures
+            auto meritNew = computeMeritCFS();
+            double delta = merit != 0.0 ? abs(merit - meritNew) / merit : 0.0;
+            if (meritNew > merit || delta < threshold) {
+                if (meritNew > merit) {
+                    merit = meritNew;
+                }
+                selectedScores.push_back(meritNew);
+            } else {
+                selectedFeatures.pop_back();
+                break;
+            }
+            if (selectedFeatures.size() == maxFeatures) {
+                break;
+            }
+        }
+        fitted = true;
+    }
+}

src/BayesNet/IWSS.h

@@ -0,0 +1,18 @@
+#ifndef IWSS_H
+#define IWSS_H
+#include <torch/torch.h>
+#include <vector>
+#include "FeatureSelect.h"
+using namespace std;
+namespace bayesnet {
+    class IWSS : public FeatureSelect {
+    public:
+        // dataset is a n+1xm tensor of integers where dataset[-1] is the y vector
+        IWSS(const torch::Tensor& samples, const vector<string>& features, const string& className, const int maxFeatures, const int classNumStates, const torch::Tensor& weights, const double threshold);
+        virtual ~IWSS() {};
+        void fit() override;
+    private:
+        double threshold = -1;
+    };
+}
+#endif

src/BayesNet/KDB.cc

@@ -1,11 +1,22 @@
 #include "KDB.h"
 
 namespace bayesnet {
-    using namespace std;
     using namespace torch;
 
-    KDB::KDB(int k, float theta) : BaseClassifier(Network()), k(k), theta(theta) {}
-    void KDB::train()
+    KDB::KDB(int k, float theta) : Classifier(Network()), k(k), theta(theta) {}
+    void KDB::setHyperparameters(nlohmann::json& hyperparameters)
+    {
+        // Check if hyperparameters are valid
+        const vector<string> validKeys = { "k", "theta" };
+        checkHyperparameters(validKeys, hyperparameters);
+        if (hyperparameters.contains("k")) {
+            k = hyperparameters["k"];
+        }
+        if (hyperparameters.contains("theta")) {
+            theta = hyperparameters["theta"];
+        }
+    }
+    void KDB::buildModel(const torch::Tensor& weights)
     {
         /*
         1. For each feature Xi, compute mutual information, I(X;C),
@@ -28,25 +39,25 @@ namespace bayesnet {
         */
         // 1. For each feature Xi, compute mutual information, I(X;C),
         // where C is the class.
-        vector <float> mi;
+        addNodes();
+        const Tensor& y = dataset.index({ -1, "..." });
+        vector<double> mi;
         for (auto i = 0; i < features.size(); i++) {
-            Tensor firstFeature = X.index({ "...", i });
-            mi.push_back(metrics.mutualInformation(firstFeature, y));
+            Tensor firstFeature = dataset.index({ i, "..." });
+            mi.push_back(metrics.mutualInformation(firstFeature, y, weights));
         }
         // 2. Compute class conditional mutual information I(Xi;XjIC), f or each
-        auto conditionalEdgeWeights = metrics.conditionalEdge();
+        auto conditionalEdgeWeights = metrics.conditionalEdge(weights);
         // 3. Let the used variable list, S, be empty.
         vector<int> S;
         // 4. Let the DAG network being constructed, BN, begin with a single
         // class node, C.
-        model.addNode(className, states[className].size());
         // 5. Repeat until S includes all domain features
         // 5.1. Select feature Xmax which is not in S and has the largest value
         // I(Xmax;C).
         auto order = argsort(mi);
         for (auto idx : order) {
             // 5.2. Add a node to BN representing Xmax.
-            model.addNode(features[idx], states[features[idx]].size());
             // 5.3. Add an arc from C to Xmax in BN.
             model.addEdge(className, features[idx]);
             // 5.4. Add m = min(lSl,/c) arcs from m distinct features Xj in S with
@@ -80,11 +91,12 @@ namespace bayesnet {
             exit_cond = num == n_edges || candidates.size(0) == 0;
         }
     }
-    vector<string> KDB::graph(string title)
+    vector<string> KDB::graph(const string& title) const
     {
+        string header{ title };
         if (title == "KDB") {
-            title += " (k=" + to_string(k) + ", theta=" + to_string(theta) + ")";
+            header += " (k=" + to_string(k) + ", theta=" + to_string(theta) + ")";
         }
-        return model.graph(title);
+        return model.graph(header);
     }
 }

src/BayesNet/KDB.h

@@ -0,0 +1,23 @@
+#ifndef KDB_H
+#define KDB_H
+#include <torch/torch.h>
+#include "Classifier.h"
+#include "bayesnetUtils.h"
+namespace bayesnet {
+    using namespace std;
+    using namespace torch;
+    class KDB : public Classifier {
+    private:
+        int k;
+        float theta;
+        void add_m_edges(int idx, vector<int>& S, Tensor& weights);
+    protected:
+        void buildModel(const torch::Tensor& weights) override;
+    public:
+        explicit KDB(int k, float theta = 0.03);
+        virtual ~KDB() {};
+        void setHyperparameters(nlohmann::json& hyperparameters) override;
+        vector<string> graph(const string& name = "KDB") const override;
+    };
+}
+#endif

src/BayesNet/KDBLd.cc

@@ -0,0 +1,30 @@
+#include "KDBLd.h"
+
+namespace bayesnet {
+    using namespace std;
+    KDBLd::KDBLd(int k) : KDB(k), Proposal(dataset, features, className) {}
+    KDBLd& KDBLd::fit(torch::Tensor& X_, torch::Tensor& y_, const vector<string>& features_, const string& className_, map<string, vector<int>>& states_)
+    {
+        checkInput(X_, y_);
+        features = features_;
+        className = className_;
+        Xf = X_;
+        y = y_;
+        // Fills vectors Xv & yv with the data from tensors X_ (discretized) & y
+        states = fit_local_discretization(y);
+        // We have discretized the input data
+        // 1st we need to fit the model to build the normal KDB structure, KDB::fit initializes the base Bayesian network
+        KDB::fit(dataset, features, className, states);
+        states = localDiscretizationProposal(states, model);
+        return *this;
+    }
+    Tensor KDBLd::predict(Tensor& X)
+    {
+        auto Xt = prepareX(X);
+        return KDB::predict(Xt);
+    }
+    vector<string> KDBLd::graph(const string& name) const
+    {
+        return KDB::graph(name);
+    }
+}

src/BayesNet/KDBLd.h

@@ -0,0 +1,19 @@
+#ifndef KDBLD_H
+#define KDBLD_H
+#include "KDB.h"
+#include "Proposal.h"
+
+namespace bayesnet {
+    using namespace std;
+    class KDBLd : public KDB, public Proposal {
+    private:
+    public:
+        explicit KDBLd(int k);
+        virtual ~KDBLd() = default;
+        KDBLd& fit(torch::Tensor& X, torch::Tensor& y, const vector<string>& features, const string& className, map<string, vector<int>>& states) override;
+        vector<string> graph(const string& name = "KDB") const override;
+        Tensor predict(Tensor& X) override;
+        static inline string version() { return "0.0.1"; };
+    };
+}
+#endif // !KDBLD_H

src/BayesNet/Mst.cc

@@ -1,5 +1,6 @@
 #include "Mst.h"
 #include <vector>
+#include <list>
 /*
     Based on the code from https://www.softwaretestinghelp.com/minimum-spanning-tree-tutorial/
 
@@ -7,9 +8,8 @@
 
 namespace bayesnet {
     using namespace std;
-    Graph::Graph(int V)
+    Graph::Graph(int V) : V(V), parent(vector<int>(V))
     {
-        parent = vector<int>(V);
         for (int i = 0; i < V; i++)
             parent[i] = i;
         G.clear();
@@ -34,10 +34,10 @@ namespace bayesnet {
     }
     void Graph::kruskal_algorithm()
     {
-        int i, uSt, vEd;
         // sort the edges ordered on decreasing weight
-        sort(G.begin(), G.end(), [](auto& left, auto& right) {return left.first > right.first;});
-        for (i = 0; i < G.size(); i++) {
+        stable_sort(G.begin(), G.end(), [](const auto& left, const auto& right) {return left.first > right.first;});
+        for (int i = 0; i < G.size(); i++) {
+            int uSt, vEd;
             uSt = find_set(G[i].second.first);
             vEd = find_set(G[i].second.second);
             if (uSt != vEd) {
@@ -56,15 +56,24 @@ namespace bayesnet {
         }
     }
 
+    void insertElement(list<int>& variables, int variable)
+    {
+        if (find(variables.begin(), variables.end(), variable) == variables.end()) {
+            variables.push_front(variable);
+        }
+    }
+
     vector<pair<int, int>> reorder(vector<pair<float, pair<int, int>>> T, int root_original)
     {
+        // Create the edges of a DAG from the MST
+        // replacing unordered_set with list because unordered_set cannot guarantee the order of the elements inserted
         auto result = vector<pair<int, int>>();
         auto visited = vector<int>();
-        auto nextVariables = unordered_set<int>();
-        nextVariables.emplace(root_original);
+        auto nextVariables = list<int>();
+        nextVariables.push_front(root_original);
         while (nextVariables.size() > 0) {
-            int root = *nextVariables.begin();
-            nextVariables.erase(nextVariables.begin());
+            int root = nextVariables.front();
+            nextVariables.pop_front();
             for (int i = 0; i < T.size(); ++i) {
                 auto [weight, edge] = T[i];
                 auto [from, to] = edge;
@@ -72,10 +81,10 @@ namespace bayesnet {
                     visited.insert(visited.begin(), i);
                     if (from == root) {
                         result.push_back({ from, to });
-                        nextVariables.emplace(to);
+                        insertElement(nextVariables, to);
                     } else {
                         result.push_back({ to, from });
-                        nextVariables.emplace(from);
+                        insertElement(nextVariables, from);
                     }
                 }
             }
@@ -95,15 +104,14 @@ namespace bayesnet {
         return result;
     }
 
-    MST::MST(vector<string>& features, Tensor& weights, int root) : features(features), weights(weights), root(root) {}
+    MST::MST(const vector<string>& features, const Tensor& weights, const int root) : features(features), weights(weights), root(root) {}
     vector<pair<int, int>> MST::maximumSpanningTree()
     {
         auto num_features = features.size();
         Graph g(num_features);
-
         // Make a complete graph
         for (int i = 0; i < num_features - 1; ++i) {
-            for (int j = i; j < num_features; ++j) {
+            for (int j = i + 1; j < num_features; ++j) {
                 g.addEdge(i, j, weights[i][j].item<float>());
             }
         }

src/BayesNet/Mst.h

@@ -10,10 +10,10 @@ namespace bayesnet {
     private:
         Tensor weights;
         vector<string> features;
-        int root;
+        int root = 0;
     public:
         MST() = default;
-        MST(vector<string>& features, Tensor& weights, int root);
+        MST(const vector<string>& features, const Tensor& weights, const int root);
         vector<pair<int, int>> maximumSpanningTree();
     };
     class Graph {
@@ -23,7 +23,7 @@ namespace bayesnet {
         vector <pair<float, pair<int, int>>> T; // vector for mst
         vector<int> parent;
     public:
-        Graph(int V);
+        explicit Graph(int V);
         void addEdge(int u, int v, float wt);
         int find_set(int i);
         void union_set(int u, int v);

src/BayesNet/Network.cc

@@ -0,0 +1,413 @@
+#include <thread>
+#include <mutex>
+#include "Network.h"
+#include "bayesnetUtils.h"
+namespace bayesnet {
+    Network::Network() : features(vector<string>()), className(""), classNumStates(0), fitted(false), laplaceSmoothing(0) {}
+    Network::Network(float maxT) : features(vector<string>()), className(""), classNumStates(0), maxThreads(maxT), fitted(false), laplaceSmoothing(0) {}
+    Network::Network(Network& other) : laplaceSmoothing(other.laplaceSmoothing), features(other.features), className(other.className), classNumStates(other.getClassNumStates()), maxThreads(other.
+        getmaxThreads()), fitted(other.fitted)
+    {
+        for (const auto& pair : other.nodes) {
+            nodes[pair.first] = std::make_unique<Node>(*pair.second);
+        }
+    }
+    void Network::initialize()
+    {
+        features = vector<string>();
+        className = "";
+        classNumStates = 0;
+        fitted = false;
+        nodes.clear();
+        samples = torch::Tensor();
+    }
+    float Network::getmaxThreads()
+    {
+        return maxThreads;
+    }
+    torch::Tensor& Network::getSamples()
+    {
+        return samples;
+    }
+    void Network::addNode(const string& name)
+    {
+        if (name == "") {
+            throw invalid_argument("Node name cannot be empty");
+        }
+        if (nodes.find(name) != nodes.end()) {
+            return;
+        }
+        if (find(features.begin(), features.end(), name) == features.end()) {
+            features.push_back(name);
+        }
+        nodes[name] = std::make_unique<Node>(name);
+    }
+    vector<string> Network::getFeatures() const
+    {
+        return features;
+    }
+    int Network::getClassNumStates() const
+    {
+        return classNumStates;
+    }
+    int Network::getStates() const
+    {
+        int result = 0;
+        for (auto& node : nodes) {
+            result += node.second->getNumStates();
+        }
+        return result;
+    }
+    string Network::getClassName() const
+    {
+        return className;
+    }
+    bool Network::isCyclic(const string& nodeId, unordered_set<string>& visited, unordered_set<string>& recStack)
+    {
+        if (visited.find(nodeId) == visited.end()) // if node hasn't been visited yet
+        {
+            visited.insert(nodeId);
+            recStack.insert(nodeId);
+            for (Node* child : nodes[nodeId]->getChildren()) {
+                if (visited.find(child->getName()) == visited.end() && isCyclic(child->getName(), visited, recStack))
+                    return true;
+                else if (recStack.find(child->getName()) != recStack.end())
+                    return true;
+            }
+        }
+        recStack.erase(nodeId); // remove node from recursion stack before function ends
+        return false;
+    }
+    void Network::addEdge(const string& parent, const string& child)
+    {
+        if (nodes.find(parent) == nodes.end()) {
+            throw invalid_argument("Parent node " + parent + " does not exist");
+        }
+        if (nodes.find(child) == nodes.end()) {
+            throw invalid_argument("Child node " + child + " does not exist");
+        }
+        // Temporarily add edge to check for cycles
+        nodes[parent]->addChild(nodes[child].get());
+        nodes[child]->addParent(nodes[parent].get());
+        unordered_set<string> visited;
+        unordered_set<string> recStack;
+        if (isCyclic(nodes[child]->getName(), visited, recStack)) // if adding this edge forms a cycle
+        {
+            // remove problematic edge
+            nodes[parent]->removeChild(nodes[child].get());
+            nodes[child]->removeParent(nodes[parent].get());
+            throw invalid_argument("Adding this edge forms a cycle in the graph.");
+        }
+    }
+    map<string, std::unique_ptr<Node>>& Network::getNodes()
+    {
+        return nodes;
+    }
+    void Network::checkFitData(int n_samples, int n_features, int n_samples_y, const vector<string>& featureNames, const string& className, const map<string, vector<int>>& states, const torch::Tensor& weights)
+    {
+        if (weights.size(0) != n_samples) {
+            throw invalid_argument("Weights (" + to_string(weights.size(0)) + ") must have the same number of elements as samples (" + to_string(n_samples) + ") in Network::fit");
+        }
+        if (n_samples != n_samples_y) {
+            throw invalid_argument("X and y must have the same number of samples in Network::fit (" + to_string(n_samples) + " != " + to_string(n_samples_y) + ")");
+        }
+        if (n_features != featureNames.size()) {
+            throw invalid_argument("X and features must have the same number of features in Network::fit (" + to_string(n_features) + " != " + to_string(featureNames.size()) + ")");
+        }
+        if (n_features != features.size() - 1) {
+            throw invalid_argument("X and local features must have the same number of features in Network::fit (" + to_string(n_features) + " != " + to_string(features.size() - 1) + ")");
+        }
+        if (find(features.begin(), features.end(), className) == features.end()) {
+            throw invalid_argument("className not found in Network::features");
+        }
+        for (auto& feature : featureNames) {
+            if (find(features.begin(), features.end(), feature) == features.end()) {
+                throw invalid_argument("Feature " + feature + " not found in Network::features");
+            }
+            if (states.find(feature) == states.end()) {
+                throw invalid_argument("Feature " + feature + " not found in states");
+            }
+        }
+    }
+    void Network::setStates(const map<string, vector<int>>& states)
+    {
+        // Set states to every Node in the network
+        for_each(features.begin(), features.end(), [this, &states](const string& feature) {
+            nodes.at(feature)->setNumStates(states.at(feature).size());
+            });
+        classNumStates = nodes.at(className)->getNumStates();
+    }
+    // X comes in nxm, where n is the number of features and m the number of samples
+    void Network::fit(const torch::Tensor& X, const torch::Tensor& y, const torch::Tensor& weights, const vector<string>& featureNames, const string& className, const map<string, vector<int>>& states)
+    {
+        checkFitData(X.size(1), X.size(0), y.size(0), featureNames, className, states, weights);
+        this->className = className;
+        Tensor ytmp = torch::transpose(y.view({ y.size(0), 1 }), 0, 1);
+        samples = torch::cat({ X , ytmp }, 0);
+        for (int i = 0; i < featureNames.size(); ++i) {
+            auto row_feature = X.index({ i, "..." });
+        }
+        completeFit(states, weights);
+    }
+    void Network::fit(const torch::Tensor& samples, const torch::Tensor& weights, const vector<string>& featureNames, const string& className, const map<string, vector<int>>& states)
+    {
+        checkFitData(samples.size(1), samples.size(0) - 1, samples.size(1), featureNames, className, states, weights);
+        this->className = className;
+        this->samples = samples;
+        completeFit(states, weights);
+    }
+    // input_data comes in nxm, where n is the number of features and m the number of samples
+    void Network::fit(const vector<vector<int>>& input_data, const vector<int>& labels, const vector<double>& weights_, const vector<string>& featureNames, const string& className, const map<string, vector<int>>& states)
+    {
+        const torch::Tensor weights = torch::tensor(weights_, torch::kFloat64);
+        checkFitData(input_data[0].size(), input_data.size(), labels.size(), featureNames, className, states, weights);
+        this->className = className;
+        // Build tensor of samples (nxm) (n+1 because of the class)
+        samples = torch::zeros({ static_cast<int>(input_data.size() + 1), static_cast<int>(input_data[0].size()) }, torch::kInt32);
+        for (int i = 0; i < featureNames.size(); ++i) {
+            samples.index_put_({ i, "..." }, torch::tensor(input_data[i], torch::kInt32));
+        }
+        samples.index_put_({ -1, "..." }, torch::tensor(labels, torch::kInt32));
+        completeFit(states, weights);
+    }
+    void Network::completeFit(const map<string, vector<int>>& states, const torch::Tensor& weights)
+    {
+        setStates(states);
+        laplaceSmoothing = 1.0 / samples.size(1); // To use in CPT computation
+        vector<thread> threads;
+        for (auto& node : nodes) {
+            threads.emplace_back([this, &node, &weights]() {
+                node.second->computeCPT(samples, features, laplaceSmoothing, weights);
+                });
+        }
+        for (auto& thread : threads) {
+            thread.join();
+        }
+        fitted = true;
+    }
+    torch::Tensor Network::predict_tensor(const torch::Tensor& samples, const bool proba)
+    {
+        if (!fitted) {
+            throw logic_error("You must call fit() before calling predict()");
+        }
+        torch::Tensor result;
+        result = torch::zeros({ samples.size(1), classNumStates }, torch::kFloat64);
+        for (int i = 0; i < samples.size(1); ++i) {
+            const Tensor sample = samples.index({ "...", i });
+            auto psample = predict_sample(sample);
+            auto temp = torch::tensor(psample, torch::kFloat64);
+            //            result.index_put_({ i, "..." }, torch::tensor(predict_sample(sample), torch::kFloat64));
+            result.index_put_({ i, "..." }, temp);
+        }
+        if (proba)
+            return result;
+        return result.argmax(1);
+    }
+    // Return mxn tensor of probabilities
+    Tensor Network::predict_proba(const Tensor& samples)
+    {
+        return predict_tensor(samples, true);
+    }
+
+    // Return mxn tensor of probabilities
+    Tensor Network::predict(const Tensor& samples)
+    {
+        return predict_tensor(samples, false);
+    }
+
+    // Return mx1 vector of predictions
+    // tsamples is nxm vector of samples
+    vector<int> Network::predict(const vector<vector<int>>& tsamples)
+    {
+        if (!fitted) {
+            throw logic_error("You must call fit() before calling predict()");
+        }
+        vector<int> predictions;
+        vector<int> sample;
+        for (int row = 0; row < tsamples[0].size(); ++row) {
+            sample.clear();
+            for (int col = 0; col < tsamples.size(); ++col) {
+                sample.push_back(tsamples[col][row]);
+            }
+            vector<double> classProbabilities = predict_sample(sample);
+            // Find the class with the maximum posterior probability
+            auto maxElem = max_element(classProbabilities.begin(), classProbabilities.end());
+            int predictedClass = distance(classProbabilities.begin(), maxElem);
+            predictions.push_back(predictedClass);
+        }
+        return predictions;
+    }
+    // Return mxn vector of probabilities
+    vector<vector<double>> Network::predict_proba(const vector<vector<int>>& tsamples)
+    {
+        if (!fitted) {
+            throw logic_error("You must call fit() before calling predict_proba()");
+        }
+        vector<vector<double>> predictions;
+        vector<int> sample;
+        for (int row = 0; row < tsamples[0].size(); ++row) {
+            sample.clear();
+            for (int col = 0; col < tsamples.size(); ++col) {
+                sample.push_back(tsamples[col][row]);
+            }
+            predictions.push_back(predict_sample(sample));
+        }
+        return predictions;
+    }
+    double Network::score(const vector<vector<int>>& tsamples, const vector<int>& labels)
+    {
+        vector<int> y_pred = predict(tsamples);
+        int correct = 0;
+        for (int i = 0; i < y_pred.size(); ++i) {
+            if (y_pred[i] == labels[i]) {
+                correct++;
+            }
+        }
+        return (double)correct / y_pred.size();
+    }
+    // Return 1xn vector of probabilities
+    vector<double> Network::predict_sample(const vector<int>& sample)
+    {
+        // Ensure the sample size is equal to the number of features
+        if (sample.size() != features.size() - 1) {
+            throw invalid_argument("Sample size (" + to_string(sample.size()) +
+                ") does not match the number of features (" + to_string(features.size() - 1) + ")");
+        }
+        map<string, int> evidence;
+        for (int i = 0; i < sample.size(); ++i) {
+            evidence[features[i]] = sample[i];
+        }
+        return exactInference(evidence);
+    }
+    // Return 1xn vector of probabilities
+    vector<double> Network::predict_sample(const Tensor& sample)
+    {
+        // Ensure the sample size is equal to the number of features
+        if (sample.size(0) != features.size() - 1) {
+            throw invalid_argument("Sample size (" + to_string(sample.size(0)) +
+                ") does not match the number of features (" + to_string(features.size() - 1) + ")");
+        }
+        map<string, int> evidence;
+        for (int i = 0; i < sample.size(0); ++i) {
+            evidence[features[i]] = sample[i].item<int>();
+        }
+        return exactInference(evidence);
+    }
+    double Network::computeFactor(map<string, int>& completeEvidence)
+    {
+        double result = 1.0;
+        for (auto& node : getNodes()) {
+            result *= node.second->getFactorValue(completeEvidence);
+        }
+        return result;
+    }
+    vector<double> Network::exactInference(map<string, int>& evidence)
+    {
+        vector<double> result(classNumStates, 0.0);
+        vector<thread> threads;
+        mutex mtx;
+        for (int i = 0; i < classNumStates; ++i) {
+            threads.emplace_back([this, &result, &evidence, i, &mtx]() {
+                auto completeEvidence = map<string, int>(evidence);
+                completeEvidence[getClassName()] = i;
+                double factor = computeFactor(completeEvidence);
+                lock_guard<mutex> lock(mtx);
+                result[i] = factor;
+                });
+        }
+        for (auto& thread : threads) {
+            thread.join();
+        }
+        // Normalize result
+        double sum = accumulate(result.begin(), result.end(), 0.0);
+        transform(result.begin(), result.end(), result.begin(), [sum](const double& value) { return value / sum; });
+        return result;
+    }
+    vector<string> Network::show() const
+    {
+        vector<string> result;
+        // Draw the network
+        for (auto& node : nodes) {
+            string line = node.first + " -> ";
+            for (auto child : node.second->getChildren()) {
+                line += child->getName() + ", ";
+            }
+            result.push_back(line);
+        }
+        return result;
+    }
+    vector<string> Network::graph(const string& title) const
+    {
+        auto output = vector<string>();
+        auto prefix = "digraph BayesNet {\nlabel=<BayesNet ";
+        auto suffix = ">\nfontsize=30\nfontcolor=blue\nlabelloc=t\nlayout=circo\n";
+        string header = prefix + title + suffix;
+        output.push_back(header);
+        for (auto& node : nodes) {
+            auto result = node.second->graph(className);
+            output.insert(output.end(), result.begin(), result.end());
+        }
+        output.push_back("}\n");
+        return output;
+    }
+    vector<pair<string, string>> Network::getEdges() const
+    {
+        auto edges = vector<pair<string, string>>();
+        for (const auto& node : nodes) {
+            auto head = node.first;
+            for (const auto& child : node.second->getChildren()) {
+                auto tail = child->getName();
+                edges.push_back({ head, tail });
+            }
+        }
+        return edges;
+    }
+    int Network::getNumEdges() const
+    {
+        return getEdges().size();
+    }
+    vector<string> Network::topological_sort()
+    {
+        /* Check if al the fathers of every node are before the node */
+        auto result = features;
+        result.erase(remove(result.begin(), result.end(), className), result.end());
+        bool ending{ false };
+        while (!ending) {
+            ending = true;
+            for (auto feature : features) {
+                auto fathers = nodes[feature]->getParents();
+                for (const auto& father : fathers) {
+                    auto fatherName = father->getName();
+                    if (fatherName == className) {
+                        continue;
+                    }
+                    // Check if father is placed before the actual feature
+                    auto it = find(result.begin(), result.end(), fatherName);
+                    if (it != result.end()) {
+                        auto it2 = find(result.begin(), result.end(), feature);
+                        if (it2 != result.end()) {
+                            if (distance(it, it2) < 0) {
+                                // if it is not, insert it before the feature
+                                result.erase(remove(result.begin(), result.end(), fatherName), result.end());
+                                result.insert(it2, fatherName);
+                                ending = false;
+                            }
+                        } else {
+                            throw logic_error("Error in topological sort because of node " + feature + " is not in result");
+                        }
+                    } else {
+                        throw logic_error("Error in topological sort because of node father " + fatherName + " is not in result");
+                    }
+                }
+            }
+        }
+        return result;
+    }
+    void Network::dump_cpt() const
+    {
+        for (auto& node : nodes) {
+            cout << "* " << node.first << ": (" << node.second->getNumStates() << ") : " << node.second->getCPT().sizes() << endl;
+            cout << node.second->getCPT() << endl;
+        }
+    }
+}

src/BayesNet/Network.h

@@ -0,0 +1,62 @@
+#ifndef NETWORK_H
+#define NETWORK_H
+#include "Node.h"
+#include <map>
+#include <vector>
+
+namespace bayesnet {
+    class Network {
+    private:
+        map<string, unique_ptr<Node>> nodes;
+        bool fitted;
+        float maxThreads = 0.95;
+        int classNumStates;
+        vector<string> features; // Including classname
+        string className;
+        double laplaceSmoothing;
+        torch::Tensor samples; // nxm tensor used to fit the model
+        bool isCyclic(const std::string&, std::unordered_set<std::string>&, std::unordered_set<std::string>&);
+        vector<double> predict_sample(const vector<int>&);
+        vector<double> predict_sample(const torch::Tensor&);
+        vector<double> exactInference(map<string, int>&);
+        double computeFactor(map<string, int>&);
+        void completeFit(const map<string, vector<int>>& states, const torch::Tensor& weights);
+        void checkFitData(int n_features, int n_samples, int n_samples_y, const vector<string>& featureNames, const string& className, const map<string, vector<int>>& states, const torch::Tensor& weights);
+        void setStates(const map<string, vector<int>>&);
+    public:
+        Network();
+        explicit Network(float);
+        explicit Network(Network&);
+        ~Network() = default;
+        torch::Tensor& getSamples();
+        float getmaxThreads();
+        void addNode(const string&);
+        void addEdge(const string&, const string&);
+        map<string, std::unique_ptr<Node>>& getNodes();
+        vector<string> getFeatures() const;
+        int getStates() const;
+        vector<pair<string, string>> getEdges() const;
+        int getNumEdges() const;
+        int getClassNumStates() const;
+        string getClassName() const;
+        /*
+        Notice: Nodes have to be inserted in the same order as they are in the dataset, i.e., first node is first column and so on.
+        */
+        void fit(const vector<vector<int>>& input_data, const vector<int>& labels, const vector<double>& weights, const vector<string>& featureNames, const string& className, const map<string, vector<int>>& states);
+        void fit(const torch::Tensor& X, const torch::Tensor& y, const torch::Tensor& weights, const vector<string>& featureNames, const string& className, const map<string, vector<int>>& states);
+        void fit(const torch::Tensor& samples, const torch::Tensor& weights, const vector<string>& featureNames, const string& className, const map<string, vector<int>>& states);
+        vector<int> predict(const vector<vector<int>>&); // Return mx1 vector of predictions
+        torch::Tensor predict(const torch::Tensor&); // Return mx1 tensor of predictions
+        torch::Tensor predict_tensor(const torch::Tensor& samples, const bool proba);
+        vector<vector<double>> predict_proba(const vector<vector<int>>&); // Return mxn vector of probabilities
+        torch::Tensor predict_proba(const torch::Tensor&); // Return mxn tensor of probabilities
+        double score(const vector<vector<int>>&, const vector<int>&);
+        vector<string> topological_sort();
+        vector<string> show() const;
+        vector<string> graph(const string& title) const; // Returns a vector of strings representing the graph in graphviz format
+        void initialize();
+        void dump_cpt() const;
+        inline string version() { return "0.2.0"; }
+    };
+}
+#endif

src/BayesNet/Node.cc

@@ -2,10 +2,18 @@
 
 namespace bayesnet {
 
-    Node::Node(const std::string& name, int numStates)
-        : name(name), numStates(numStates), cpTable(torch::Tensor()), parents(vector<Node*>()), children(vector<Node*>())
+    Node::Node(const std::string& name)
+        : name(name), numStates(0), cpTable(torch::Tensor()), parents(vector<Node*>()), children(vector<Node*>())
     {
     }
+    void Node::clear()
+    {
+        parents.clear();
+        children.clear();
+        cpTable = torch::Tensor();
+        dimensions.clear();
+        numStates = 0;
+    }
     string Node::getName() const
     {
         return name;
@@ -76,47 +84,52 @@ namespace bayesnet {
         }
         return result;
     }
-    void Node::computeCPT(map<string, vector<int>>& dataset, const int laplaceSmoothing)
+    void Node::computeCPT(const torch::Tensor& dataset, const vector<string>& features, const double laplaceSmoothing, const torch::Tensor& weights)
     {
+        dimensions.clear();
         // Get dimensions of the CPT
         dimensions.push_back(numStates);
-        for (auto father : getParents()) {
-            dimensions.push_back(father->getNumStates());
-        }
-        auto length = dimensions.size();
+        transform(parents.begin(), parents.end(), back_inserter(dimensions), [](const auto& parent) { return parent->getNumStates(); });
+
         // Create a tensor of zeros with the dimensions of the CPT
         cpTable = torch::zeros(dimensions, torch::kFloat) + laplaceSmoothing;
         // Fill table with counts
-        for (int n_sample = 0; n_sample < dataset[name].size(); ++n_sample) {
-            torch::List<c10::optional<torch::Tensor>> coordinates;
-            coordinates.push_back(torch::tensor(dataset[name][n_sample]));
-            for (auto father : getParents()) {
-                coordinates.push_back(torch::tensor(dataset[father->getName()][n_sample]));
+        auto pos = find(features.begin(), features.end(), name);
+        if (pos == features.end()) {
+            throw logic_error("Feature " + name + " not found in dataset");
+        }
+        int name_index = pos - features.begin();
+        for (int n_sample = 0; n_sample < dataset.size(1); ++n_sample) {
+            c10::List<c10::optional<at::Tensor>> coordinates;
+            coordinates.push_back(dataset.index({ name_index, n_sample }));
+            for (auto parent : parents) {
+                pos = find(features.begin(), features.end(), parent->getName());
+                if (pos == features.end()) {
+                    throw logic_error("Feature parent " + parent->getName() + " not found in dataset");
+                }
+                int parent_index = pos - features.begin();
+                coordinates.push_back(dataset.index({ parent_index, n_sample }));
             }
             // Increment the count of the corresponding coordinate
-            cpTable.index_put_({ coordinates }, cpTable.index({ coordinates }) + 1);
+            cpTable.index_put_({ coordinates }, cpTable.index({ coordinates }) + weights.index({ n_sample }).item<double>());
         }
         // Normalize the counts
         cpTable = cpTable / cpTable.sum(0);
     }
     float Node::getFactorValue(map<string, int>& evidence)
     {
-        torch::List<c10::optional<torch::Tensor>> coordinates;
+        c10::List<c10::optional<at::Tensor>> coordinates;
         // following predetermined order of indices in the cpTable (see Node.h)
-        coordinates.push_back(torch::tensor(evidence[name]));
-        for (auto parent : getParents()) {
-            coordinates.push_back(torch::tensor(evidence[parent->getName()]));
-        }
+        coordinates.push_back(at::tensor(evidence[name]));
+        transform(parents.begin(), parents.end(), back_inserter(coordinates), [&evidence](const auto& parent) { return at::tensor(evidence[parent->getName()]); });
         return cpTable.index({ coordinates }).item<float>();
     }
-    vector<string> Node::graph(string className)
+    vector<string> Node::graph(const string& className)
     {
         auto output = vector<string>();
         auto suffix = name == className ? ", fontcolor=red, fillcolor=lightblue, style=filled " : "";
         output.push_back(name + " [shape=circle" + suffix + "] \n");
-        for (auto& child : children) {
-            output.push_back(name + " -> " + child->getName());
-        }
+        transform(children.begin(), children.end(), back_inserter(output), [this](const auto& child) { return name + " -> " + child->getName(); });
         return output;
     }
 }

src/BayesNet/Node.h

@@ -14,9 +14,10 @@ namespace bayesnet {
         int numStates; // number of states of the variable
         torch::Tensor cpTable; // Order of indices is 0-> node variable, 1-> 1st parent, 2-> 2nd parent, ...
         vector<int64_t> dimensions; // dimensions of the cpTable
-    public:
         vector<pair<string, string>> combinations(const vector<string>&);
-        Node(const std::string&, int);
+    public:
+        explicit Node(const string&);
+        void clear();
         void addParent(Node*);
         void addChild(Node*);
         void removeParent(Node*);
@@ -25,11 +26,11 @@ namespace bayesnet {
         vector<Node*>& getParents();
         vector<Node*>& getChildren();
         torch::Tensor& getCPT();
-        void computeCPT(map<string, vector<int>>&, const int);
+        void computeCPT(const torch::Tensor& dataset, const vector<string>& features, const double laplaceSmoothing, const torch::Tensor& weights);
         int getNumStates() const;
         void setNumStates(int);
         unsigned minFill();
-        vector<string> graph(string clasName); // Returns a vector of strings representing the graph in graphviz format
+        vector<string> graph(const string& clasName); // Returns a vector of strings representing the graph in graphviz format
         float getFactorValue(map<string, int>&);
     };
 }

src/BayesNet/Proposal.cc

@@ -0,0 +1,110 @@
+#include "Proposal.h"
+#include "ArffFiles.h"
+
+namespace bayesnet {
+    Proposal::Proposal(torch::Tensor& dataset_, vector<string>& features_, string& className_) : pDataset(dataset_), pFeatures(features_), pClassName(className_) {}
+    Proposal::~Proposal()
+    {
+        for (auto& [key, value] : discretizers) {
+            delete value;
+        }
+    }
+    void Proposal::checkInput(const torch::Tensor& X, const torch::Tensor& y)
+    {
+        if (!torch::is_floating_point(X)) {
+            throw std::invalid_argument("X must be a floating point tensor");
+        }
+        if (torch::is_floating_point(y)) {
+            throw std::invalid_argument("y must be an integer tensor");
+        }
+    }
+    map<string, vector<int>> Proposal::localDiscretizationProposal(const map<string, vector<int>>& oldStates, Network& model)
+    {
+        // order of local discretization is important. no good 0, 1, 2...
+        // although we rediscretize features after the local discretization of every feature
+        auto order = model.topological_sort();
+        auto& nodes = model.getNodes();
+        map<string, vector<int>> states = oldStates;
+        vector<int> indicesToReDiscretize;
+        bool upgrade = false; // Flag to check if we need to upgrade the model
+        for (auto feature : order) {
+            auto nodeParents = nodes[feature]->getParents();
+            if (nodeParents.size() < 2) continue; // Only has class as parent
+            upgrade = true;
+            int index = find(pFeatures.begin(), pFeatures.end(), feature) - pFeatures.begin();
+            indicesToReDiscretize.push_back(index); // We need to re-discretize this feature
+            vector<string> parents;
+            transform(nodeParents.begin(), nodeParents.end(), back_inserter(parents), [](const auto& p) { return p->getName(); });
+            // Remove class as parent as it will be added later
+            parents.erase(remove(parents.begin(), parents.end(), pClassName), parents.end());
+            // Get the indices of the parents
+            vector<int> indices;
+            indices.push_back(-1); // Add class index
+            transform(parents.begin(), parents.end(), back_inserter(indices), [&](const auto& p) {return find(pFeatures.begin(), pFeatures.end(), p) - pFeatures.begin(); });
+            // Now we fit the discretizer of the feature, conditioned on its parents and the class i.e. discretizer.fit(X[index], X[indices] + y)
+            vector<string> yJoinParents(Xf.size(1));
+            for (auto idx : indices) {
+                for (int i = 0; i < Xf.size(1); ++i) {
+                    yJoinParents[i] += to_string(pDataset.index({ idx, i }).item<int>());
+                }
+            }
+            auto arff = ArffFiles();
+            auto yxv = arff.factorize(yJoinParents);
+            auto xvf_ptr = Xf.index({ index }).data_ptr<float>();
+            auto xvf = vector<mdlp::precision_t>(xvf_ptr, xvf_ptr + Xf.size(1));
+            discretizers[feature]->fit(xvf, yxv);
+        }
+        if (upgrade) {
+            // Discretize again X (only the affected indices) with the new fitted discretizers
+            for (auto index : indicesToReDiscretize) {
+                auto Xt_ptr = Xf.index({ index }).data_ptr<float>();
+                auto Xt = vector<float>(Xt_ptr, Xt_ptr + Xf.size(1));
+                pDataset.index_put_({ index, "..." }, torch::tensor(discretizers[pFeatures[index]]->transform(Xt)));
+                auto xStates = vector<int>(discretizers[pFeatures[index]]->getCutPoints().size() + 1);
+                iota(xStates.begin(), xStates.end(), 0);
+                //Update new states of the feature/node
+                states[pFeatures[index]] = xStates;
+            }
+            const torch::Tensor weights = torch::full({ pDataset.size(1) }, 1.0 / pDataset.size(1), torch::kDouble);
+            model.fit(pDataset, weights, pFeatures, pClassName, states);
+        }
+        return states;
+    }
+    map<string, vector<int>> Proposal::fit_local_discretization(const torch::Tensor& y)
+    {
+        // Discretize the continuous input data and build pDataset (Classifier::dataset)
+        int m = Xf.size(1);
+        int n = Xf.size(0);
+        map<string, vector<int>> states;
+        pDataset = torch::zeros({ n + 1, m }, kInt32);
+        auto yv = vector<int>(y.data_ptr<int>(), y.data_ptr<int>() + y.size(0));
+        // discretize input data by feature(row)
+        for (auto i = 0; i < pFeatures.size(); ++i) {
+            auto* discretizer = new mdlp::CPPFImdlp();
+            auto Xt_ptr = Xf.index({ i }).data_ptr<float>();
+            auto Xt = vector<float>(Xt_ptr, Xt_ptr + Xf.size(1));
+            discretizer->fit(Xt, yv);
+            pDataset.index_put_({ i, "..." }, torch::tensor(discretizer->transform(Xt)));
+            auto xStates = vector<int>(discretizer->getCutPoints().size() + 1);
+            iota(xStates.begin(), xStates.end(), 0);
+            states[pFeatures[i]] = xStates;
+            discretizers[pFeatures[i]] = discretizer;
+        }
+        int n_classes = torch::max(y).item<int>() + 1;
+        auto yStates = vector<int>(n_classes);
+        iota(yStates.begin(), yStates.end(), 0);
+        states[pClassName] = yStates;
+        pDataset.index_put_({ n, "..." }, y);
+        return states;
+    }
+    torch::Tensor Proposal::prepareX(torch::Tensor& X)
+    {
+        auto Xtd = torch::zeros_like(X, torch::kInt32);
+        for (int i = 0; i < X.size(0); ++i) {
+            auto Xt = vector<float>(X[i].data_ptr<float>(), X[i].data_ptr<float>() + X.size(1));
+            auto Xd = discretizers[pFeatures[i]]->transform(Xt);
+            Xtd.index_put_({ i }, torch::tensor(Xd, torch::kInt32));
+        }
+        return Xtd;
+    }
+}

src/BayesNet/Proposal.h

@@ -0,0 +1,30 @@
+#ifndef PROPOSAL_H
+#define PROPOSAL_H
+#include <string>
+#include <map>
+#include <torch/torch.h>
+#include "Network.h"
+#include "CPPFImdlp.h"
+#include "Classifier.h"
+
+namespace bayesnet {
+    class Proposal {
+    public:
+        Proposal(torch::Tensor& pDataset, vector<string>& features_, string& className_);
+        virtual ~Proposal();
+    protected:
+        void checkInput(const torch::Tensor& X, const torch::Tensor& y);
+        torch::Tensor prepareX(torch::Tensor& X);
+        map<string, vector<int>> localDiscretizationProposal(const map<string, vector<int>>& states, Network& model);
+        map<string, vector<int>> fit_local_discretization(const torch::Tensor& y);
+        torch::Tensor Xf; // X continuous nxm tensor
+        torch::Tensor y; // y discrete nx1 tensor
+        map<string, mdlp::CPPFImdlp*> discretizers;
+    private:
+        torch::Tensor& pDataset; // (n+1)xm tensor
+        vector<string>& pFeatures;
+        string& pClassName;
+    };
+}
+
+#endif  

src/BayesNet/SPODE.cc

@@ -2,9 +2,9 @@
 
 namespace bayesnet {
 
-    SPODE::SPODE(int root) : BaseClassifier(Network()), root(root) {}
+    SPODE::SPODE(int root) : Classifier(Network()), root(root) {}
 
-    void SPODE::train()
+    void SPODE::buildModel(const torch::Tensor& weights)
     {
         // 0. Add all nodes to the model
         addNodes();
@@ -17,7 +17,7 @@ namespace bayesnet {
             }
         }
     }
-    vector<string> SPODE::graph(string name )
+    vector<string> SPODE::graph(const string& name) const
     {
         return model.graph(name);
     }

src/BayesNet/SPODE.h

@@ -0,0 +1,17 @@
+#ifndef SPODE_H
+#define SPODE_H
+#include "Classifier.h"
+
+namespace bayesnet {
+    class SPODE : public Classifier {
+    private:
+        int root;
+    protected:
+        void buildModel(const torch::Tensor& weights) override;
+    public:
+        explicit SPODE(int root);
+        virtual ~SPODE() {};
+        vector<string> graph(const string& name = "SPODE") const override;
+    };
+}
+#endif

src/BayesNet/SPODELd.cc

@@ -0,0 +1,48 @@
+#include "SPODELd.h"
+
+namespace bayesnet {
+    using namespace std;
+    SPODELd::SPODELd(int root) : SPODE(root), Proposal(dataset, features, className) {}
+    SPODELd& SPODELd::fit(torch::Tensor& X_, torch::Tensor& y_, const vector<string>& features_, const string& className_, map<string, vector<int>>& states_)
+    {
+        checkInput(X_, y_);
+        features = features_;
+        className = className_;
+        Xf = X_;
+        y = y_;
+        // Fills vectors Xv & yv with the data from tensors X_ (discretized) & y
+        states = fit_local_discretization(y);
+        // We have discretized the input data
+        // 1st we need to fit the model to build the normal SPODE structure, SPODE::fit initializes the base Bayesian network
+        SPODE::fit(dataset, features, className, states);
+        states = localDiscretizationProposal(states, model);
+        return *this;
+    }
+    SPODELd& SPODELd::fit(torch::Tensor& dataset, const vector<string>& features_, const string& className_, map<string, vector<int>>& states_)
+    {
+        if (!torch::is_floating_point(dataset)) {
+            throw std::runtime_error("Dataset must be a floating point tensor");
+        }
+        Xf = dataset.index({ torch::indexing::Slice(0, dataset.size(0) - 1), "..." }).clone();
+        y = dataset.index({ -1, "..." }).clone();
+        features = features_;
+        className = className_;
+        // Fills vectors Xv & yv with the data from tensors X_ (discretized) & y
+        states = fit_local_discretization(y);
+        // We have discretized the input data
+        // 1st we need to fit the model to build the normal SPODE structure, SPODE::fit initializes the base Bayesian network
+        SPODE::fit(dataset, features, className, states);
+        states = localDiscretizationProposal(states, model);
+        return *this;
+    }
+
+    Tensor SPODELd::predict(Tensor& X)
+    {
+        auto Xt = prepareX(X);
+        return SPODE::predict(Xt);
+    }
+    vector<string> SPODELd::graph(const string& name) const
+    {
+        return SPODE::graph(name);
+    }
+}

src/BayesNet/SPODELd.h

@@ -0,0 +1,19 @@
+#ifndef SPODELD_H
+#define SPODELD_H
+#include "SPODE.h"
+#include "Proposal.h"
+
+namespace bayesnet {
+    using namespace std;
+    class SPODELd : public SPODE, public Proposal {
+    public:
+        explicit SPODELd(int root);
+        virtual ~SPODELd() = default;
+        SPODELd& fit(torch::Tensor& X, torch::Tensor& y, const vector<string>& features, const string& className, map<string, vector<int>>& states) override;
+        SPODELd& fit(torch::Tensor& dataset, const vector<string>& features, const string& className, map<string, vector<int>>& states) override;
+        vector<string> graph(const string& name = "SPODE") const override;
+        Tensor predict(Tensor& X) override;
+        static inline string version() { return "0.0.1"; };
+    };
+}
+#endif // !SPODELD_H

src/BayesNet/TAN.cc

@@ -1,30 +1,29 @@
 #include "TAN.h"
 
 namespace bayesnet {
-    using namespace std;
     using namespace torch;
 
-    TAN::TAN() : BaseClassifier(Network()) {}
+    TAN::TAN() : Classifier(Network()) {}
 
-    void TAN::train()
+    void TAN::buildModel(const torch::Tensor& weights)
     {
         // 0. Add all nodes to the model
         addNodes();
         // 1. Compute mutual information between each feature and the class and set the root node
         // as the highest mutual information with the class
         auto mi = vector <pair<int, float >>();
-        Tensor class_dataset = dataset.index({ "...", -1 });
+        Tensor class_dataset = dataset.index({ -1, "..." });
         for (int i = 0; i < static_cast<int>(features.size()); ++i) {
-            Tensor feature_dataset = dataset.index({ "...", i });
-            auto mi_value = metrics.mutualInformation(class_dataset, feature_dataset);
+            Tensor feature_dataset = dataset.index({ i, "..." });
+            auto mi_value = metrics.mutualInformation(class_dataset, feature_dataset, weights);
             mi.push_back({ i, mi_value });
         }
-        sort(mi.begin(), mi.end(), [](auto& left, auto& right) {return left.second < right.second;});
+        sort(mi.begin(), mi.end(), [](const auto& left, const auto& right) {return left.second < right.second;});
         auto root = mi[mi.size() - 1].first;
         // 2. Compute mutual information between each feature and the class
-        auto weights = metrics.conditionalEdge();
+        auto weights_matrix = metrics.conditionalEdge(weights);
         // 3. Compute the maximum spanning tree
-        auto mst = metrics.maximumSpanningTree(features, weights, root);
+        auto mst = metrics.maximumSpanningTree(features, weights_matrix, root);
         // 4. Add edges from the maximum spanning tree to the model
         for (auto i = 0; i < mst.size(); ++i) {
             auto [from, to] = mst[i];
@@ -35,7 +34,7 @@ namespace bayesnet {
             model.addEdge(className, feature);
         }
     }
-    vector<string> TAN::graph(string title)
+    vector<string> TAN::graph(const string& title) const
     {
         return model.graph(title);
     }

src/BayesNet/TAN.h

@@ -0,0 +1,16 @@
+#ifndef TAN_H
+#define TAN_H
+#include "Classifier.h"
+namespace bayesnet {
+    using namespace std;
+    class TAN : public Classifier {
+    private:
+    protected:
+        void buildModel(const torch::Tensor& weights) override;
+    public:
+        TAN();
+        virtual ~TAN() {};
+        vector<string> graph(const string& name = "TAN") const override;
+    };
+}
+#endif

src/BayesNet/TANLd.cc

@@ -0,0 +1,31 @@
+#include "TANLd.h"
+
+namespace bayesnet {
+    using namespace std;
+    TANLd::TANLd() : TAN(), Proposal(dataset, features, className) {}
+    TANLd& TANLd::fit(torch::Tensor& X_, torch::Tensor& y_, const vector<string>& features_, const string& className_, map<string, vector<int>>& states_)
+    {
+        checkInput(X_, y_);
+        features = features_;
+        className = className_;
+        Xf = X_;
+        y = y_;
+        // Fills vectors Xv & yv with the data from tensors X_ (discretized) & y
+        states = fit_local_discretization(y);
+        // We have discretized the input data
+        // 1st we need to fit the model to build the normal TAN structure, TAN::fit initializes the base Bayesian network
+        TAN::fit(dataset, features, className, states);
+        states = localDiscretizationProposal(states, model);
+        return *this;
+
+    }
+    Tensor TANLd::predict(Tensor& X)
+    {
+        auto Xt = prepareX(X);
+        return TAN::predict(Xt);
+    }
+    vector<string> TANLd::graph(const string& name) const
+    {
+        return TAN::graph(name);
+    }
+}

src/BayesNet/TANLd.h

@@ -0,0 +1,19 @@
+#ifndef TANLD_H
+#define TANLD_H
+#include "TAN.h"
+#include "Proposal.h"
+
+namespace bayesnet {
+    using namespace std;
+    class TANLd : public TAN, public Proposal {
+    private:
+    public:
+        TANLd();
+        virtual ~TANLd() = default;
+        TANLd& fit(torch::Tensor& X, torch::Tensor& y, const vector<string>& features, const string& className, map<string, vector<int>>& states) override;
+        vector<string> graph(const string& name = "TAN") const override;
+        Tensor predict(Tensor& X) override;
+        static inline string version() { return "0.0.1"; };
+    };
+}
+#endif // !TANLD_H

src/BayesNet/bayesnetUtils.cc

@@ -0,0 +1,27 @@
+
+#include "bayesnetUtils.h"
+namespace bayesnet {
+    using namespace std;
+    using namespace torch;
+    // Return the indices in descending order
+    vector<int> argsort(vector<double>& nums)
+    {
+        int n = nums.size();
+        vector<int> indices(n);
+        iota(indices.begin(), indices.end(), 0);
+        sort(indices.begin(), indices.end(), [&nums](int i, int j) {return nums[i] > nums[j];});
+        return indices;
+    }
+    vector<vector<int>> tensorToVector(Tensor& tensor)
+    {
+        // convert mxn tensor to nxm vector
+        vector<vector<int>> result;
+        // Iterate over cols
+        for (int i = 0; i < tensor.size(1); ++i) {
+            auto col_tensor = tensor.index({ "...", i });
+            auto col = vector<int>(col_tensor.data_ptr<int>(), col_tensor.data_ptr<int>() + tensor.size(0));
+            result.push_back(col);
+        }
+        return result;
+    }
+}

src/BayesNet/bayesnetUtils.h

@@ -0,0 +1,11 @@
+#ifndef BAYESNET_UTILS_H
+#define BAYESNET_UTILS_H
+#include <torch/torch.h>
+#include <vector>
+namespace bayesnet {
+    using namespace std;
+    using namespace torch;
+    vector<int> argsort(vector<double>& nums);
+    vector<vector<int>> tensorToVector(Tensor& tensor);
+}
+#endif //BAYESNET_UTILS_H

src/CMakeLists.txt

@@ -1,2 +0,0 @@
-add_library(BayesNet utils.cc Network.cc Node.cc Metrics.cc BaseClassifier.cc KDB.cc TAN.cc SPODE.cc Ensemble.cc AODE.cc Mst.cc)
-target_link_libraries(BayesNet "${TORCH_LIBRARIES}")

src/Ensemble.cc

@@ -1,112 +0,0 @@
-#include "Ensemble.h"
-
-namespace bayesnet {
-    using namespace std;
-    using namespace torch;
-
-    Ensemble::Ensemble() : m(0), n(0), n_models(0), metrics(Metrics()), fitted(false) {}
-    Ensemble& Ensemble::build(vector<string>& features, string className, map<string, vector<int>>& states)
-    {
-        dataset = cat({ X, y.view({y.size(0), 1}) }, 1);
-        this->features = features;
-        this->className = className;
-        this->states = states;
-        auto n_classes = states[className].size();
-        metrics = Metrics(dataset, features, className, n_classes);
-        // Build models
-        train();
-        // Train models
-        n_models = models.size();
-        for (auto i = 0; i < n_models; ++i) {
-            models[i]->fit(Xv, yv, features, className, states);
-        }
-        fitted = true;
-        return *this;
-    }
-    Ensemble& Ensemble::fit(vector<vector<int>>& X, vector<int>& y, vector<string>& features, string className, map<string, vector<int>>& states)
-    {
-        this->X = torch::zeros({ static_cast<int64_t>(X[0].size()), static_cast<int64_t>(X.size()) }, kInt64);
-        Xv = X;
-        for (int i = 0; i < X.size(); ++i) {
-            this->X.index_put_({ "...", i }, torch::tensor(X[i], kInt64));
-        }
-        this->y = torch::tensor(y, kInt64);
-        yv = y;
-        return build(features, className, states);
-    }
-    Tensor Ensemble::predict(Tensor& X)
-    {
-        if (!fitted) {
-            throw logic_error("Ensemble has not been fitted");
-        }
-        Tensor y_pred = torch::zeros({ X.size(0), n_models }, kInt64);
-        for (auto i = 0; i < n_models; ++i) {
-            y_pred.index_put_({ "...", i }, models[i]->predict(X));
-        }
-        return torch::tensor(voting(y_pred));
-    }
-    vector<int> Ensemble::voting(Tensor& y_pred)
-    {
-        auto y_pred_ = y_pred.accessor<int64_t, 2>();
-        vector<int> y_pred_final;
-        for (int i = 0; i < y_pred.size(0); ++i) {
-            vector<float> votes(states[className].size(), 0);
-            for (int j = 0; j < y_pred.size(1); ++j) {
-                votes[y_pred_[i][j]] += 1;
-            }
-            auto indices = argsort(votes);
-            y_pred_final.push_back(indices[0]);
-        }
-        return y_pred_final;
-    }
-    vector<int> Ensemble::predict(vector<vector<int>>& X)
-    {
-        if (!fitted) {
-            throw logic_error("Ensemble has not been fitted");
-        }
-        long m_ = X[0].size();
-        long n_ = X.size();
-        vector<vector<int>> Xd(n_, vector<int>(m_, 0));
-        for (auto i = 0; i < n_; i++) {
-            Xd[i] = vector<int>(X[i].begin(), X[i].end());
-        }
-        Tensor y_pred = torch::zeros({ m_, n_models }, kInt64);
-        for (auto i = 0; i < n_models; ++i) {
-            y_pred.index_put_({ "...", i }, torch::tensor(models[i]->predict(Xd), kInt64));
-        }
-        return voting(y_pred);
-    }
-    float Ensemble::score(vector<vector<int>>& X, vector<int>& y)
-    {
-        if (!fitted) {
-            throw logic_error("Ensemble has not been fitted");
-        }
-        auto y_pred = predict(X);
-        int correct = 0;
-        for (int i = 0; i < y_pred.size(); ++i) {
-            if (y_pred[i] == y[i]) {
-                correct++;
-            }
-        }
-        return (double)correct / y_pred.size();
-
-    }
-    vector<string> Ensemble::show()
-    {
-        auto result = vector<string>();
-        for (auto i = 0; i < n_models; ++i) {
-            auto res = models[i]->show();
-            result.insert(result.end(), res.begin(), res.end());
-        }
-        return result;
-    }
-    vector<string> Ensemble::graph(string title)
-    {
-        auto result = vector<string>();
-        for (auto i = 0; i < n_models; ++i) {
-            auto res = models[i]->graph(title + "_" + to_string(i));
-            result.insert(result.end(), res.begin(), res.end());
-        }
-        return result;
-    }
-}

src/Ensemble.h

@@ -1,42 +0,0 @@
-#ifndef ENSEMBLE_H
-#define ENSEMBLE_H
-#include <torch/torch.h>
-#include "BaseClassifier.h"
-#include "Metrics.hpp"
-#include "utils.h"
-using namespace std;
-using namespace torch;
-
-namespace bayesnet {
-    class Ensemble {
-    private:
-        bool fitted;
-        long n_models;
-        Ensemble& build(vector<string>& features, string className, map<string, vector<int>>& states);
-    protected:
-        vector<unique_ptr<BaseClassifier>> models;
-        int m, n; // m: number of samples, n: number of features
-        Tensor X;
-        vector<vector<int>> Xv;
-        Tensor y;
-        vector<int> yv;
-        Tensor dataset;
-        Metrics metrics;
-        vector<string> features;
-        string className;
-        map<string, vector<int>> states;
-        void virtual train() = 0;
-        vector<int> voting(Tensor& y_pred);
-    public:
-        Ensemble();
-        virtual ~Ensemble() = default;
-        Ensemble& fit(vector<vector<int>>& X, vector<int>& y, vector<string>& features, string className, map<string, vector<int>>& states);
-        Tensor predict(Tensor& X);
-        vector<int> predict(vector<vector<int>>& X);
-        float score(Tensor& X, Tensor& y);
-        float score(vector<vector<int>>& X, vector<int>& y);
-        vector<string> show();
-        vector<string> graph(string title);
-    };
-}
-#endif

src/KDB.h

@@ -1,20 +0,0 @@
-#ifndef KDB_H
-#define KDB_H
-#include "BaseClassifier.h"
-#include "utils.h"
-namespace bayesnet {
-    using namespace std;
-    using namespace torch;
-    class KDB : public BaseClassifier {
-    private:
-        int k;
-        float theta;
-        void add_m_edges(int idx, vector<int>& S, Tensor& weights);
-    protected:
-        void train() override;
-    public:
-        KDB(int k, float theta = 0.03);
-        vector<string> graph(string name = "KDB") override;
-    };
-}
-#endif

src/Metrics.cc

@@ -1,131 +0,0 @@
-#include "Metrics.hpp"
-#include "Mst.h"
-using namespace std;
-namespace bayesnet {
-    Metrics::Metrics(torch::Tensor& samples, vector<string>& features, string& className, int classNumStates)
-        : samples(samples)
-        , features(features)
-        , className(className)
-        , classNumStates(classNumStates)
-    {
-    }
-    Metrics::Metrics(const vector<vector<int>>& vsamples, const vector<int>& labels, const vector<string>& features, const string& className, const int classNumStates)
-        : features(features)
-        , className(className)
-        , classNumStates(classNumStates)
-    {
-        samples = torch::zeros({ static_cast<int64_t>(vsamples[0].size()), static_cast<int64_t>(vsamples.size() + 1) }, torch::kInt64);
-        for (int i = 0; i < vsamples.size(); ++i) {
-            samples.index_put_({ "...", i }, torch::tensor(vsamples[i], torch::kInt64));
-        }
-        samples.index_put_({ "...", -1 }, torch::tensor(labels, torch::kInt64));
-    }
-    vector<pair<string, string>> Metrics::doCombinations(const vector<string>& source)
-    {
-        vector<pair<string, string>> result;
-        for (int i = 0; i < source.size(); ++i) {
-            string temp = source[i];
-            for (int j = i + 1; j < source.size(); ++j) {
-                result.push_back({ temp, source[j] });
-            }
-        }
-        return result;
-    }
-    torch::Tensor Metrics::conditionalEdge()
-    {
-        auto result = vector<double>();
-        auto source = vector<string>(features);
-        source.push_back(className);
-        auto combinations = doCombinations(source);
-        // Compute class prior
-        auto margin = torch::zeros({ classNumStates });
-        for (int value = 0; value < classNumStates; ++value) {
-            auto mask = samples.index({ "...", -1 }) == value;
-            margin[value] = mask.sum().item<float>() / samples.sizes()[0];
-        }
-        for (auto [first, second] : combinations) {
-            int64_t index_first = find(features.begin(), features.end(), first) - features.begin();
-            int64_t index_second = find(features.begin(), features.end(), second) - features.begin();
-            double accumulated = 0;
-            for (int value = 0; value < classNumStates; ++value) {
-                auto mask = samples.index({ "...", -1 }) == value;
-                auto first_dataset = samples.index({ mask, index_first });
-                auto second_dataset = samples.index({ mask, index_second });
-                auto mi = mutualInformation(first_dataset, second_dataset);
-                auto pb = margin[value].item<float>();
-                accumulated += pb * mi;
-            }
-            result.push_back(accumulated);
-        }
-        long n_vars = source.size();
-        auto matrix = torch::zeros({ n_vars, n_vars });
-        auto indices = torch::triu_indices(n_vars, n_vars, 1);
-        for (auto i = 0; i < result.size(); ++i) {
-            auto x = indices[0][i];
-            auto y = indices[1][i];
-            matrix[x][y] = result[i];
-            matrix[y][x] = result[i];
-        }
-        return matrix;
-    }
-    vector<float> Metrics::conditionalEdgeWeights()
-    {
-        auto matrix = conditionalEdge();
-        std::vector<float> v(matrix.data_ptr<float>(), matrix.data_ptr<float>() + matrix.numel());
-        return v;
-    }
-    double Metrics::entropy(torch::Tensor& feature)
-    {
-        torch::Tensor counts = feature.bincount();
-        int totalWeight = counts.sum().item<int>();
-        torch::Tensor probs = counts.to(torch::kFloat) / totalWeight;
-        torch::Tensor logProbs = torch::log(probs);
-        torch::Tensor entropy = -probs * logProbs;
-        return entropy.nansum().item<double>();
-    }
-    // H(Y|X) = sum_{x in X} p(x) H(Y|X=x)
-    double Metrics::conditionalEntropy(torch::Tensor& firstFeature, torch::Tensor& secondFeature)
-    {
-        int numSamples = firstFeature.sizes()[0];
-        torch::Tensor featureCounts = secondFeature.bincount();
-        unordered_map<int, unordered_map<int, double>> jointCounts;
-        double totalWeight = 0;
-        for (auto i = 0; i < numSamples; i++) {
-            jointCounts[secondFeature[i].item<int>()][firstFeature[i].item<int>()] += 1;
-            totalWeight += 1;
-        }
-        if (totalWeight == 0)
-            throw invalid_argument("Total weight should not be zero");
-        double entropyValue = 0;
-        for (int value = 0; value < featureCounts.sizes()[0]; ++value) {
-            double p_f = featureCounts[value].item<double>() / totalWeight;
-            double entropy_f = 0;
-            for (auto& [label, jointCount] : jointCounts[value]) {
-                double p_l_f = jointCount / featureCounts[value].item<double>();
-                if (p_l_f > 0) {
-                    entropy_f -= p_l_f * log(p_l_f);
-                } else {
-                    entropy_f = 0;
-                }
-            }
-            entropyValue += p_f * entropy_f;
-        }
-        return entropyValue;
-    }
-    // I(X;Y) = H(Y) - H(Y|X)
-    double Metrics::mutualInformation(torch::Tensor& firstFeature, torch::Tensor& secondFeature)
-    {
-        return entropy(firstFeature) - conditionalEntropy(firstFeature, secondFeature);
-    }
-    /*
-    Compute the maximum spanning tree considering the weights as distances
-    and the indices of the weights as nodes of this square matrix using
-    Kruskal algorithm
-    */
-    vector<pair<int, int>> Metrics::maximumSpanningTree(vector<string> features, Tensor& weights, int root)
-    {
-        auto result = vector<pair<int, int>>();
-        auto mst = MST(features, weights, root);
-        return mst.maximumSpanningTree();
-    }
-}

src/Metrics.hpp

@@ -1,28 +0,0 @@
-#ifndef BAYESNET_METRICS_H
-#define BAYESNET_METRICS_H
-#include <torch/torch.h>
-#include <vector>
-#include <string>
-namespace bayesnet {
-    using namespace std;
-    using namespace torch;
-    class Metrics {
-    private:
-        Tensor samples;
-        vector<string> features;
-        string className;
-        int classNumStates;
-    public:
-        Metrics() = default;
-        Metrics(Tensor&, vector<string>&, string&, int);
-        Metrics(const vector<vector<int>>&, const vector<int>&, const vector<string>&, const string&, const int);
-        double entropy(Tensor&);
-        double conditionalEntropy(Tensor&, Tensor&);
-        double mutualInformation(Tensor&, Tensor&);
-        vector<float> conditionalEdgeWeights();
-        Tensor conditionalEdge();
-        vector<pair<string, string>> doCombinations(const vector<string>&);
-        vector<pair<int, int>> maximumSpanningTree(vector<string> features, Tensor& weights, int root);
-    };
-}
-#endif

src/Network.cc

@@ -1,276 +0,0 @@
-#include <thread>
-#include <mutex>
-#include "Network.h"
-namespace bayesnet {
-    Network::Network() : laplaceSmoothing(1), features(vector<string>()), className(""), classNumStates(0), maxThreads(0.8), fitted(false) {}
-    Network::Network(float maxT) : laplaceSmoothing(1), features(vector<string>()), className(""), classNumStates(0), maxThreads(maxT), fitted(false) {}
-    Network::Network(float maxT, int smoothing) : laplaceSmoothing(smoothing), features(vector<string>()), className(""), classNumStates(0), maxThreads(maxT), fitted(false) {}
-    Network::Network(Network& other) : laplaceSmoothing(other.laplaceSmoothing), features(other.features), className(other.className), classNumStates(other.getClassNumStates()), maxThreads(other.getmaxThreads()), fitted(other.fitted)
-    {
-        for (auto& pair : other.nodes) {
-            nodes[pair.first] = make_unique<Node>(*pair.second);
-        }
-    }
-    float Network::getmaxThreads()
-    {
-        return maxThreads;
-    }
-    torch::Tensor& Network::getSamples()
-    {
-        return samples;
-    }
-    void Network::addNode(string name, int numStates)
-    {
-        if (nodes.find(name) != nodes.end()) {
-            // if node exists update its number of states
-            nodes[name]->setNumStates(numStates);
-            return;
-        }
-        nodes[name] = make_unique<Node>(name, numStates);
-    }
-    vector<string> Network::getFeatures()
-    {
-        return features;
-    }
-    int Network::getClassNumStates()
-    {
-        return classNumStates;
-    }
-    int Network::getStates()
-    {
-        int result = 0;
-        for (auto& node : nodes) {
-            result += node.second->getNumStates();
-        }
-        return result;
-    }
-    string Network::getClassName()
-    {
-        return className;
-    }
-    bool Network::isCyclic(const string& nodeId, unordered_set<string>& visited, unordered_set<string>& recStack)
-    {
-        if (visited.find(nodeId) == visited.end()) // if node hasn't been visited yet
-        {
-            visited.insert(nodeId);
-            recStack.insert(nodeId);
-            for (Node* child : nodes[nodeId]->getChildren()) {
-                if (visited.find(child->getName()) == visited.end() && isCyclic(child->getName(), visited, recStack))
-                    return true;
-                else if (recStack.find(child->getName()) != recStack.end())
-                    return true;
-            }
-        }
-        recStack.erase(nodeId); // remove node from recursion stack before function ends
-        return false;
-    }
-    void Network::addEdge(const string parent, const string child)
-    {
-        if (nodes.find(parent) == nodes.end()) {
-            throw invalid_argument("Parent node " + parent + " does not exist");
-        }
-        if (nodes.find(child) == nodes.end()) {
-            throw invalid_argument("Child node " + child + " does not exist");
-        }
-        // Temporarily add edge to check for cycles
-        nodes[parent]->addChild(nodes[child].get());
-        nodes[child]->addParent(nodes[parent].get());
-        unordered_set<string> visited;
-        unordered_set<string> recStack;
-        if (isCyclic(nodes[child]->getName(), visited, recStack)) // if adding this edge forms a cycle
-        {
-            // remove problematic edge
-            nodes[parent]->removeChild(nodes[child].get());
-            nodes[child]->removeParent(nodes[parent].get());
-            throw invalid_argument("Adding this edge forms a cycle in the graph.");
-        }
-
-    }
-    map<string, std::unique_ptr<Node>>& Network::getNodes()
-    {
-        return nodes;
-    }
-    void Network::fit(const vector<vector<int>>& input_data, const vector<int>& labels, const vector<string>& featureNames, const string& className)
-    {
-        features = featureNames;
-        this->className = className;
-        dataset.clear();
-
-        // Build dataset & tensor of samples
-        samples = torch::zeros({ static_cast<int64_t>(input_data[0].size()), static_cast<int64_t>(input_data.size() + 1) }, torch::kInt64);
-        for (int i = 0; i < featureNames.size(); ++i) {
-            dataset[featureNames[i]] = input_data[i];
-            samples.index_put_({ "...", i }, torch::tensor(input_data[i], torch::kInt64));
-        }
-        dataset[className] = labels;
-        samples.index_put_({ "...", -1 }, torch::tensor(labels, torch::kInt64));
-        classNumStates = *max_element(labels.begin(), labels.end()) + 1;
-        int maxThreadsRunning = static_cast<int>(std::thread::hardware_concurrency() * maxThreads);
-        if (maxThreadsRunning < 1) {
-            maxThreadsRunning = 1;
-        }
-        vector<thread> threads;
-        mutex mtx;
-        condition_variable cv;
-        int activeThreads = 0;
-        int nextNodeIndex = 0;
-
-        while (nextNodeIndex < nodes.size()) {
-            unique_lock<mutex> lock(mtx);
-            cv.wait(lock, [&activeThreads, &maxThreadsRunning]() { return activeThreads < maxThreadsRunning; });
-
-            if (nextNodeIndex >= nodes.size()) {
-                break;  // No more work remaining
-            }
-
-            threads.emplace_back([this, &nextNodeIndex, &mtx, &cv, &activeThreads]() {
-                while (true) {
-                    unique_lock<mutex> lock(mtx);
-                    if (nextNodeIndex >= nodes.size()) {
-                        break;  // No more work remaining
-                    }
-                    auto& pair = *std::next(nodes.begin(), nextNodeIndex);
-                    ++nextNodeIndex;
-                    lock.unlock();
-
-                    pair.second->computeCPT(dataset, laplaceSmoothing);
-                    lock.lock();
-                    nodes[pair.first] = std::move(pair.second);
-                    lock.unlock();
-                }
-                lock_guard<mutex> lock(mtx);
-                --activeThreads;
-                cv.notify_one();
-                });
-
-            ++activeThreads;
-        }
-        for (auto& thread : threads) {
-            thread.join();
-        }
-        fitted = true;
-    }
-
-    vector<int> Network::predict(const vector<vector<int>>& tsamples)
-    {
-        if (!fitted) {
-            throw logic_error("You must call fit() before calling predict()");
-        }
-        vector<int> predictions;
-        vector<int> sample;
-        for (int row = 0; row < tsamples[0].size(); ++row) {
-            sample.clear();
-            for (int col = 0; col < tsamples.size(); ++col) {
-                sample.push_back(tsamples[col][row]);
-            }
-            vector<double> classProbabilities = predict_sample(sample);
-            // Find the class with the maximum posterior probability
-            auto maxElem = max_element(classProbabilities.begin(), classProbabilities.end());
-            int predictedClass = distance(classProbabilities.begin(), maxElem);
-            predictions.push_back(predictedClass);
-        }
-        return predictions;
-    }
-    vector<vector<double>> Network::predict_proba(const vector<vector<int>>& tsamples)
-    {
-        if (!fitted) {
-            throw logic_error("You must call fit() before calling predict_proba()");
-        }
-        vector<vector<double>> predictions;
-        vector<int> sample;
-        for (int row = 0; row < tsamples[0].size(); ++row) {
-            sample.clear();
-            for (int col = 0; col < tsamples.size(); ++col) {
-                sample.push_back(tsamples[col][row]);
-            }
-            predictions.push_back(predict_sample(sample));
-        }
-        return predictions;
-    }
-    double Network::score(const vector<vector<int>>& tsamples, const vector<int>& labels)
-    {
-        vector<int> y_pred = predict(tsamples);
-        int correct = 0;
-        for (int i = 0; i < y_pred.size(); ++i) {
-            if (y_pred[i] == labels[i]) {
-                correct++;
-            }
-        }
-        return (double)correct / y_pred.size();
-    }
-    vector<double> Network::predict_sample(const vector<int>& sample)
-    {
-        // Ensure the sample size is equal to the number of features
-        if (sample.size() != features.size()) {
-            throw invalid_argument("Sample size (" + to_string(sample.size()) +
-                ") does not match the number of features (" + to_string(features.size()) + ")");
-        }
-        map<string, int> evidence;
-        for (int i = 0; i < sample.size(); ++i) {
-            evidence[features[i]] = sample[i];
-        }
-        return exactInference(evidence);
-
-    }
-    double Network::computeFactor(map<string, int>& completeEvidence)
-    {
-        double result = 1.0;
-        for (auto& node : getNodes()) {
-            result *= node.second->getFactorValue(completeEvidence);
-        }
-        return result;
-    }
-    vector<double> Network::exactInference(map<string, int>& evidence)
-    {
-        vector<double> result(classNumStates, 0.0);
-        vector<thread> threads;
-        mutex mtx;
-        for (int i = 0; i < classNumStates; ++i) {
-            threads.emplace_back([this, &result, &evidence, i, &mtx]() {
-                auto completeEvidence = map<string, int>(evidence);
-                completeEvidence[getClassName()] = i;
-                double factor = computeFactor(completeEvidence);
-                lock_guard<mutex> lock(mtx);
-                result[i] = factor;
-                });
-        }
-        for (auto& thread : threads) {
-            thread.join();
-        }
-
-        // Normalize result
-        double sum = accumulate(result.begin(), result.end(), 0.0);
-        for (double& value : result) {
-            value /= sum;
-        }
-        return result;
-    }
-    vector<string> Network::show()
-    {
-        vector<string> result;
-        // Draw the network
-        for (auto& node : nodes) {
-            string line = node.first + " -> ";
-            for (auto child : node.second->getChildren()) {
-                line += child->getName() + ", ";
-            }
-            result.push_back(line);
-        }
-        return result;
-    }
-    vector<string> Network::graph(string title)
-    {
-        auto output = vector<string>();
-        auto prefix = "digraph BayesNet {\nlabel=<BayesNet ";
-        auto suffix = ">\nfontsize=30\nfontcolor=blue\nlabelloc=t\nlayout=circo\n";
-        string header = prefix + title + suffix;
-        output.push_back(header);
-        for (auto& node : nodes) {
-            auto result = node.second->graph(className);
-            output.insert(output.end(), result.begin(), result.end());
-        }
-        output.push_back("}\n");
-        return output;
-    }
-
-}

src/Network.h

@@ -1,52 +0,0 @@
-#ifndef NETWORK_H
-#define NETWORK_H
-#include "Node.h"
-#include <map>
-#include <vector>
-
-namespace bayesnet {
-    class Network {
-    private:
-        map<string, std::unique_ptr<Node>> nodes;
-        map<string, vector<int>> dataset;
-        bool fitted;
-        float maxThreads;
-        int classNumStates;
-        vector<string> features;
-        string className;
-        int laplaceSmoothing;
-        torch::Tensor samples;
-        bool isCyclic(const std::string&, std::unordered_set<std::string>&, std::unordered_set<std::string>&);
-        vector<double> predict_sample(const vector<int>&);
-        vector<double> exactInference(map<string, int>&);
-        double computeFactor(map<string, int>&);
-        double mutual_info(torch::Tensor&, torch::Tensor&);
-        double entropy(torch::Tensor&);
-        double conditionalEntropy(torch::Tensor&, torch::Tensor&);
-        double mutualInformation(torch::Tensor&, torch::Tensor&);
-    public:
-        Network();
-        Network(float, int);
-        Network(float);
-        Network(Network&);
-        torch::Tensor& getSamples();
-        float getmaxThreads();
-        void addNode(string, int);
-        void addEdge(const string, const string);
-        map<string, std::unique_ptr<Node>>& getNodes();
-        vector<string> getFeatures();
-        int getStates();
-        int getClassNumStates();
-        string getClassName();
-        void fit(const vector<vector<int>>&, const vector<int>&, const vector<string>&, const string&);
-        vector<int> predict(const vector<vector<int>>&);
-        //Computes the conditional edge weight of variable index u and v conditioned on class_node
-        torch::Tensor conditionalEdgeWeight();
-        vector<vector<double>> predict_proba(const vector<vector<int>>&);
-        double score(const vector<vector<int>>&, const vector<int>&);
-        vector<string> show();
-        vector<string> graph(string title); // Returns a vector of strings representing the graph in graphviz format
-        inline string version() { return "0.1.0"; }
-    };
-}
-#endif

src/Platform/BestResults.cc

@@ -0,0 +1,300 @@
+#include <filesystem>
+#include <set>
+#include <fstream>
+#include <iostream>
+#include <sstream>
+#include "BestResults.h"
+#include "Result.h"
+#include "Colors.h"
+#include "Statistics.h"
+#include "BestResultsExcel.h"
+#include "CLocale.h"
+
+
+namespace fs = std::filesystem;
+// function ftime_to_string, Code taken from 
+// https://stackoverflow.com/a/58237530/1389271
+template <typename TP>
+std::string ftime_to_string(TP tp)
+{
+    using namespace std::chrono;
+    auto sctp = time_point_cast<system_clock::duration>(tp - TP::clock::now()
+        + system_clock::now());
+    auto tt = system_clock::to_time_t(sctp);
+    std::tm* gmt = std::gmtime(&tt);
+    std::stringstream buffer;
+    buffer << std::put_time(gmt, "%Y-%m-%d %H:%M");
+    return buffer.str();
+}
+namespace platform {
+
+    string BestResults::build()
+    {
+        auto files = loadResultFiles();
+        if (files.size() == 0) {
+            cerr << Colors::MAGENTA() << "No result files were found!" << Colors::RESET() << endl;
+            exit(1);
+        }
+        json bests;
+        for (const auto& file : files) {
+            auto result = Result(path, file);
+            auto data = result.load();
+            for (auto const& item : data.at("results")) {
+                bool update = false;
+                // Check if results file contains only one dataset
+                auto datasetName = item.at("dataset").get<string>();
+                if (bests.contains(datasetName)) {
+                    if (item.at("score").get<double>() > bests[datasetName].at(0).get<double>()) {
+                        update = true;
+                    }
+                } else {
+                    update = true;
+                }
+                if (update) {
+                    bests[datasetName] = { item.at("score").get<double>(), item.at("hyperparameters"), file };
+                }
+            }
+        }
+        string bestFileName = path + bestResultFile();
+        if (FILE* fileTest = fopen(bestFileName.c_str(), "r")) {
+            fclose(fileTest);
+            cout << Colors::MAGENTA() << "File " << bestFileName << " already exists and it shall be overwritten." << Colors::RESET() << endl;
+        }
+        ofstream file(bestFileName);
+        file << bests;
+        file.close();
+        return bestFileName;
+    }
+
+    string BestResults::bestResultFile()
+    {
+        return "best_results_" + score + "_" + model + ".json";
+    }
+
+    pair<string, string> getModelScore(string name)
+    {
+        // results_accuracy_BoostAODE_MacBookpro16_2023-09-06_12:27:00_1.json
+        int i = 0;
+        auto pos = name.find("_");
+        auto pos2 = name.find("_", pos + 1);
+        string score = name.substr(pos + 1, pos2 - pos - 1);
+        pos = name.find("_", pos2 + 1);
+        string model = name.substr(pos2 + 1, pos - pos2 - 1);
+        return { model, score };
+    }
+
+    vector<string> BestResults::loadResultFiles()
+    {
+        vector<string> files;
+        using std::filesystem::directory_iterator;
+        string fileModel, fileScore;
+        for (const auto& file : directory_iterator(path)) {
+            auto fileName = file.path().filename().string();
+            if (fileName.find(".json") != string::npos && fileName.find("results_") == 0) {
+                tie(fileModel, fileScore) = getModelScore(fileName);
+                if (score == fileScore && (model == fileModel || model == "any")) {
+                    files.push_back(fileName);
+                }
+            }
+        }
+        return files;
+    }
+
+    json BestResults::loadFile(const string& fileName)
+    {
+        ifstream resultData(fileName);
+        if (resultData.is_open()) {
+            json data = json::parse(resultData);
+            return data;
+        }
+        throw invalid_argument("Unable to open result file. [" + fileName + "]");
+    }
+    vector<string> BestResults::getModels()
+    {
+        set<string> models;
+        vector<string> result;
+        auto files = loadResultFiles();
+        if (files.size() == 0) {
+            cerr << Colors::MAGENTA() << "No result files were found!" << Colors::RESET() << endl;
+            exit(1);
+        }
+        string fileModel, fileScore;
+        for (const auto& file : files) {
+            // extract the model from the file name
+            tie(fileModel, fileScore) = getModelScore(file);
+            // add the model to the vector of models
+            models.insert(fileModel);
+        }
+        result = vector<string>(models.begin(), models.end());
+        return result;
+    }
+    vector<string> BestResults::getDatasets(json table)
+    {
+        vector<string> datasets;
+        for (const auto& dataset : table.items()) {
+            datasets.push_back(dataset.key());
+        }
+        return datasets;
+    }
+
+    void BestResults::buildAll()
+    {
+        auto models = getModels();
+        for (const auto& model : models) {
+            cout << "Building best results for model: " << model << endl;
+            this->model = model;
+            build();
+        }
+        model = "any";
+    }
+
+    void BestResults::reportSingle()
+    {
+        string bestFileName = path + bestResultFile();
+        if (FILE* fileTest = fopen(bestFileName.c_str(), "r")) {
+            fclose(fileTest);
+        } else {
+            cerr << Colors::MAGENTA() << "File " << bestFileName << " doesn't exist." << Colors::RESET() << endl;
+            exit(1);
+        }
+        auto temp = ConfigLocale();
+        auto date = ftime_to_string(filesystem::last_write_time(bestFileName));
+        auto data = loadFile(bestFileName);
+        auto datasets = getDatasets(data);
+        int maxDatasetName = (*max_element(datasets.begin(), datasets.end(), [](const string& a, const string& b) { return a.size() < b.size(); })).size();
+        cout << Colors::GREEN() << "Best results for " << model << " and " << score << " as of " << date << endl;
+        cout << "--------------------------------------------------------" << endl;
+        cout << Colors::GREEN() << " #  " << setw(maxDatasetName + 1) << left << string("Dataset") << "Score       File                                                               Hyperparameters" << endl;
+        cout << "=== " << string(maxDatasetName, '=') << " =========== ================================================================== ================================================= " << endl;
+        auto i = 0;
+        bool odd = true;
+        for (auto const& item : data.items()) {
+            auto color = odd ? Colors::BLUE() : Colors::CYAN();
+            cout << color << setw(3) << fixed << right << i++ << " ";
+            cout << setw(maxDatasetName) << left << item.key() << " ";
+            cout << setw(11) << setprecision(9) << fixed << item.value().at(0).get<double>() << " ";
+            cout << setw(66) << item.value().at(2).get<string>() << " ";
+            cout << item.value().at(1) << " ";
+            cout << endl;
+            odd = !odd;
+        }
+    }
+    json BestResults::buildTableResults(vector<string> models)
+    {
+        json table;
+        auto maxDate = filesystem::file_time_type::max();
+        for (const auto& model : models) {
+            this->model = model;
+            string bestFileName = path + bestResultFile();
+            if (FILE* fileTest = fopen(bestFileName.c_str(), "r")) {
+                fclose(fileTest);
+            } else {
+                cerr << Colors::MAGENTA() << "File " << bestFileName << " doesn't exist." << Colors::RESET() << endl;
+                exit(1);
+            }
+            auto dateWrite = filesystem::last_write_time(bestFileName);
+            if (dateWrite < maxDate) {
+                maxDate = dateWrite;
+            }
+            auto data = loadFile(bestFileName);
+            table[model] = data;
+        }
+        table["dateTable"] = ftime_to_string(maxDate);
+        return table;
+    }
+
+    void BestResults::printTableResults(vector<string> models, json table)
+    {
+        cout << Colors::GREEN() << "Best results for " << score << " as of " << table.at("dateTable").get<string>() << endl;
+        cout << "------------------------------------------------" << endl;
+        cout << Colors::GREEN() << " #  " << setw(maxDatasetName + 1) << left << string("Dataset");
+        for (const auto& model : models) {
+            cout << setw(maxModelName) << left << model << " ";
+        }
+        cout << endl;
+        cout << "=== " << string(maxDatasetName, '=') << " ";
+        for (const auto& model : models) {
+            cout << string(maxModelName, '=') << " ";
+        }
+        cout << endl;
+        auto i = 0;
+        bool odd = true;
+        map<string, double> totals;
+        int nDatasets = table.begin().value().size();
+        for (const auto& model : models) {
+            totals[model] = 0.0;
+        }
+        auto datasets = getDatasets(table.begin().value());
+        for (auto const& dataset : datasets) {
+            auto color = odd ? Colors::BLUE() : Colors::CYAN();
+            cout << color << setw(3) << fixed << right << i++ << " ";
+            cout << setw(maxDatasetName) << left << dataset << " ";
+            double maxValue = 0;
+            // Find out the max value for this dataset
+            for (const auto& model : models) {
+                double value = table[model].at(dataset).at(0).get<double>();
+                if (value > maxValue) {
+                    maxValue = value;
+                }
+            }
+            // Print the row with red colors on max values
+            for (const auto& model : models) {
+                string efectiveColor = color;
+                double value = table[model].at(dataset).at(0).get<double>();
+                if (value == maxValue) {
+                    efectiveColor = Colors::RED();
+                }
+                totals[model] += value;
+                cout << efectiveColor << setw(maxModelName) << setprecision(maxModelName - 2) << fixed << value << " ";
+            }
+            cout << endl;
+            odd = !odd;
+        }
+        cout << Colors::GREEN() << "=== " << string(maxDatasetName, '=') << " ";
+        for (const auto& model : models) {
+            cout << string(maxModelName, '=') << " ";
+        }
+        cout << endl;
+        cout << Colors::GREEN() << setw(5 + maxDatasetName) << "    Totals...................";
+        double max = 0.0;
+        for (const auto& total : totals) {
+            if (total.second > max) {
+                max = total.second;
+            }
+        }
+        for (const auto& model : models) {
+            string efectiveColor = Colors::GREEN();
+            if (totals[model] == max) {
+                efectiveColor = Colors::RED();
+            }
+            cout << efectiveColor << right << setw(maxModelName) << setprecision(maxModelName - 4) << fixed << totals[model] << " ";
+        }
+        cout << endl;
+    }
+    void BestResults::reportAll(bool excel)
+    {
+        auto models = getModels();
+        // Build the table of results
+        json table = buildTableResults(models);
+        vector<string> datasets = getDatasets(table.begin().value());
+        maxModelName = (*max_element(models.begin(), models.end(), [](const string& a, const string& b) { return a.size() < b.size(); })).size();
+        maxModelName = max(12, maxModelName);
+        maxDatasetName = (*max_element(datasets.begin(), datasets.end(), [](const string& a, const string& b) { return a.size() < b.size(); })).size();
+        maxDatasetName = max(25, maxDatasetName);
+        // Print the table of results
+        printTableResults(models, table);
+        // Compute the Friedman test
+        map<string, map<string, float>> ranksModels;
+        if (friedman) {
+            Statistics stats(models, datasets, table, significance);
+            auto result = stats.friedmanTest();
+            stats.postHocHolmTest(result);
+            ranksModels = stats.getRanks();
+        }
+        if (excel) {
+            BestResultsExcel excel(score, models, datasets, table, ranksModels, friedman, significance);
+            excel.build();
+            cout << Colors::YELLOW() << "** Excel file generated: " << excel.getFileName() << Colors::RESET() << endl;
+        }
+    }
+}

src/Platform/BestResults.h

@@ -0,0 +1,32 @@
+#ifndef BESTRESULTS_H
+#define BESTRESULTS_H
+#include <string>
+#include <nlohmann/json.hpp>
+using namespace std;
+using json = nlohmann::json;
+namespace platform {
+    class BestResults {
+    public:
+        explicit BestResults(const string& path, const string& score, const string& model, bool friedman, double significance = 0.05) : path(path), score(score), model(model), friedman(friedman), significance(significance) {}
+        string build();
+        void reportSingle();
+        void reportAll(bool excel);
+        void buildAll();
+    private:
+        vector<string> getModels();
+        vector<string> getDatasets(json table);
+        vector<string> loadResultFiles();
+        json buildTableResults(vector<string> models);
+        void printTableResults(vector<string> models, json table);
+        string bestResultFile();
+        json loadFile(const string& fileName);
+        string path;
+        string score;
+        string model;
+        bool friedman;
+        double significance;
+        int maxModelName = 0;
+        int maxDatasetName = 0;
+    };
+}
+#endif //BESTRESULTS_H

src/Platform/BestResultsExcel.cc

@@ -0,0 +1,177 @@
+#include <sstream>
+#include "BestResultsExcel.h"
+#include "Paths.h"
+#include "Statistics.h"
+
+namespace platform {
+    BestResultsExcel::BestResultsExcel(const string& score, const vector<string>& models, const vector<string>& datasets, const json& table, const map<string, map<string, float>>& ranksModels, bool friedman, double significance) :
+        score(score), models(models), datasets(datasets), table(table), ranksModels(ranksModels), friedman(friedman), significance(significance)
+    {
+        workbook = workbook_new((Paths::excel() + fileName).c_str());
+        worksheet = workbook_add_worksheet(workbook, "Best Results");
+        setProperties("Best Results");
+        createFormats();
+        int maxModelName = (*max_element(models.begin(), models.end(), [](const string& a, const string& b) { return a.size() < b.size(); })).size();
+        modelNameSize = max(modelNameSize, maxModelName);
+        int maxDatasetName = (*max_element(datasets.begin(), datasets.end(), [](const string& a, const string& b) { return a.size() < b.size(); })).size();
+        datasetNameSize = max(datasetNameSize, maxDatasetName);
+        formatColumns();
+    }
+    BestResultsExcel::~BestResultsExcel()
+    {
+        workbook_close(workbook);
+    }
+    void BestResultsExcel::formatColumns()
+    {
+        worksheet_freeze_panes(worksheet, 4, 2);
+        vector<int> columns_sizes = { 5, datasetNameSize };
+        for (int i = 0; i < models.size(); ++i) {
+            columns_sizes.push_back(modelNameSize);
+        }
+        for (int i = 0; i < columns_sizes.size(); ++i) {
+            worksheet_set_column(worksheet, i, i, columns_sizes.at(i), NULL);
+        }
+    }
+    void BestResultsExcel::build()
+    {
+        // Create Sheet with scores
+        header(false);
+        body(false);
+        footer(false);
+        if (friedman) {
+            // Create Sheet with ranks
+            worksheet = workbook_add_worksheet(workbook, "Ranks");
+            formatColumns();
+            header(true);
+            body(true);
+            footer(true);
+            // Create Sheet with Friedman Test
+            doFriedman();
+        }
+    }
+    string BestResultsExcel::getFileName()
+    {
+        return Paths::excel() + fileName;
+    }
+    void BestResultsExcel::header(bool ranks)
+    {
+        row = 0;
+        string message = ranks ? "Ranks for score " + score : "Best results for " + score;
+        worksheet_merge_range(worksheet, 0, 0, 0, 1 + models.size(), message.c_str(), styles["headerFirst"]);
+        // Body header
+        row = 3;
+        int col = 1;
+        writeString(row, 0, "Nº", "bodyHeader");
+        writeString(row, 1, "Dataset", "bodyHeader");
+        for (const auto& model : models) {
+            writeString(row, ++col, model.c_str(), "bodyHeader");
+        }
+    }
+    void BestResultsExcel::body(bool ranks)
+    {
+        row = 4;
+        int i = 0;
+        json origin = table.begin().value();
+        for (auto const& item : origin.items()) {
+            writeInt(row, 0, i++, "ints");
+            writeString(row, 1, item.key().c_str(), "text");
+            int col = 1;
+            for (const auto& model : models) {
+                double value = ranks ? ranksModels[item.key()][model] : table[model].at(item.key()).at(0).get<double>();
+                writeDouble(row, ++col, value, "result");
+            }
+            ++row;
+        }
+    }
+    void BestResultsExcel::footer(bool ranks)
+    {
+        // Set Totals
+        writeString(row, 1, "Total", "bodyHeader");
+        int col = 1;
+        for (const auto& model : models) {
+            stringstream oss;
+            oss << "=sum(indirect(address(" << 5 << "," << col + 2 << ")):indirect(address(" << row << "," << col + 2 << ")))";
+            worksheet_write_formula(worksheet, row, ++col, oss.str().c_str(), styles["bodyHeader_odd"]);
+        }
+        if (ranks) {
+            row++;
+            writeString(row, 1, "Average ranks", "bodyHeader");
+            int col = 1;
+            for (const auto& model : models) {
+                stringstream oss;
+                oss << "=sum(indirect(address(" << 5 << "," << col + 2 << ")):indirect(address(" << row - 1 << "," << col + 2 << ")))/" << datasets.size();
+                worksheet_write_formula(worksheet, row, ++col, oss.str().c_str(), styles["bodyHeader_odd"]);
+            }
+        }
+    }
+    void BestResultsExcel::doFriedman()
+    {
+        worksheet = workbook_add_worksheet(workbook, "Friedman");
+        vector<int> columns_sizes = { 5, datasetNameSize };
+        for (int i = 0; i < models.size(); ++i) {
+            columns_sizes.push_back(modelNameSize);
+        }
+        for (int i = 0; i < columns_sizes.size(); ++i) {
+            worksheet_set_column(worksheet, i, i, columns_sizes.at(i), NULL);
+        }
+        worksheet_merge_range(worksheet, 0, 0, 0, 1 + models.size(), "Friedman Test", styles["headerFirst"]);
+        row = 2;
+        Statistics stats(models, datasets, table, significance, false);
+        auto result = stats.friedmanTest();
+        stats.postHocHolmTest(result);
+        auto friedmanResult = stats.getFriedmanResult();
+        auto holmResult = stats.getHolmResult();
+        worksheet_merge_range(worksheet, row, 0, row, 1 + models.size(), "Null hypothesis: H0 'There is no significant differences between all the classifiers.'", styles["headerSmall"]);
+        row += 2;
+        writeString(row, 1, "Friedman Q", "bodyHeader");
+        writeDouble(row, 2, friedmanResult.statistic, "bodyHeader");
+        row++;
+        writeString(row, 1, "Critical χ2 value", "bodyHeader");
+        writeDouble(row, 2, friedmanResult.criticalValue, "bodyHeader");
+        row++;
+        writeString(row, 1, "p-value", "bodyHeader");
+        writeDouble(row, 2, friedmanResult.pvalue, "bodyHeader");
+        writeString(row, 3, friedmanResult.reject ? "<" : ">", "bodyHeader");
+        writeDouble(row, 4, significance, "bodyHeader");
+        writeString(row, 5, friedmanResult.reject ? "Reject H0" : "Accept H0", "bodyHeader");
+        row += 3;
+        worksheet_merge_range(worksheet, row, 0, row, 1 + models.size(), "Holm Test", styles["headerFirst"]);
+        row += 2;
+        worksheet_merge_range(worksheet, row, 0, row, 1 + models.size(), "Null hypothesis: H0 'There is no significant differences between the control model and the other models.'", styles["headerSmall"]);
+        row += 2;
+        string controlModel = "Control Model: " + holmResult.model;
+        worksheet_merge_range(worksheet, row, 1, row, 7, controlModel.c_str(), styles["bodyHeader_odd"]);
+        row++;
+        writeString(row, 1, "Model", "bodyHeader");
+        writeString(row, 2, "p-value", "bodyHeader");
+        writeString(row, 3, "Rank", "bodyHeader");
+        writeString(row, 4, "Win", "bodyHeader");
+        writeString(row, 5, "Tie", "bodyHeader");
+        writeString(row, 6, "Loss", "bodyHeader");
+        writeString(row, 7, "Reject H0", "bodyHeader");
+        row++;
+        bool first = true;
+        for (const auto& item : holmResult.holmLines) {
+            writeString(row, 1, item.model, "text");
+            if (first) {
+                // Control model info
+                first = false;
+                writeString(row, 2, "", "text");
+                writeDouble(row, 3, item.rank, "result");
+                writeString(row, 4, "", "text");
+                writeString(row, 5, "", "text");
+                writeString(row, 6, "", "text");
+                writeString(row, 7, "", "textCentered");
+            } else {
+                // Rest of the models info
+                writeDouble(row, 2, item.pvalue, "result");
+                writeDouble(row, 3, item.rank, "result");
+                writeInt(row, 4, item.wtl.win, "ints");
+                writeInt(row, 5, item.wtl.tie, "ints");
+                writeInt(row, 6, item.wtl.loss, "ints");
+                writeString(row, 7, item.reject ? "Yes" : "No", "textCentered");
+            }
+            row++;
+        }
+    }
+}

src/Platform/BestResultsExcel.h

@@ -0,0 +1,37 @@
+#ifndef BESTRESULTS_EXCEL_H
+#define BESTRESULTS_EXCEL_H
+#include "ExcelFile.h"
+#include <vector>
+#include <map>
+#include <nlohmann/json.hpp>
+
+using namespace std;
+using json = nlohmann::json;
+
+namespace platform {
+
+    class BestResultsExcel : ExcelFile {
+    public:
+        BestResultsExcel(const string& score, const vector<string>& models, const vector<string>& datasets, const json& table, const map<string, map<string, float>>& ranks, bool friedman, double significance);
+        ~BestResultsExcel();
+        void build();
+        string getFileName();
+    private:
+        void header(bool ranks);
+        void body(bool ranks);
+        void footer(bool ranks);
+        void formatColumns();
+        void doFriedman();
+        const string fileName = "BestResults.xlsx";
+        string score;
+        vector<string> models;
+        vector<string> datasets;
+        json table;
+        map<string, map<string, float>> ranksModels;
+        bool friedman;
+        double significance;
+        int modelNameSize = 12; // Min size of the column
+        int datasetNameSize = 25; // Min size of the column
+    };
+}
+#endif //BESTRESULTS_EXCEL_H

src/Platform/BestScore.h

@@ -0,0 +1,10 @@
+#ifndef BESTSCORE_H
+#define BESTSCORE_H
+#include <string>
+class BestScore {
+public:
+    static std::string title() { return "STree_default (linear-ovo)"; }
+    static double score() { return 22.109799; }
+    static std::string scoreName() { return "accuracy"; }
+};
+#endif

src/Platform/CLocale.h

@@ -0,0 +1,24 @@
+#ifndef LOCALE_H
+#define LOCALE_H
+#include <locale>
+#include <iostream>
+#include <sstream>
+#include <string>
+using namespace std;
+namespace platform {
+    struct separation : numpunct<char> {
+        char do_decimal_point() const { return ','; }
+        char do_thousands_sep() const { return '.'; }
+        string do_grouping() const { return "\03"; }
+    };
+    class ConfigLocale {
+    public:
+        explicit ConfigLocale()
+        {
+            locale mylocale(cout.getloc(), new separation);
+            locale::global(mylocale);
+            cout.imbue(mylocale);
+        }
+    };
+}
+#endif 

src/Platform/CMakeLists.txt

@@ -0,0 +1,17 @@
+include_directories(${BayesNet_SOURCE_DIR}/src/BayesNet)
+include_directories(${BayesNet_SOURCE_DIR}/src/Platform)
+include_directories(${BayesNet_SOURCE_DIR}/lib/Files)
+include_directories(${BayesNet_SOURCE_DIR}/lib/mdlp)
+include_directories(${BayesNet_SOURCE_DIR}/lib/argparse/include)
+include_directories(${BayesNet_SOURCE_DIR}/lib/json/include)
+include_directories(${BayesNet_SOURCE_DIR}/lib/libxlsxwriter/include)
+add_executable(b_main main.cc Folding.cc Experiment.cc Datasets.cc Dataset.cc Models.cc ReportConsole.cc ReportBase.cc)
+add_executable(b_manage manage.cc Results.cc Result.cc ReportConsole.cc ReportExcel.cc ReportBase.cc Datasets.cc Dataset.cc ExcelFile.cc)
+add_executable(b_list list.cc Datasets.cc Dataset.cc)
+add_executable(b_best best.cc BestResults.cc Result.cc Statistics.cc BestResultsExcel.cc ExcelFile.cc)
+add_executable(testx testx.cpp Datasets.cc Dataset.cc Folding.cc )
+target_link_libraries(b_main BayesNet ArffFiles mdlp "${TORCH_LIBRARIES}")
+target_link_libraries(b_manage "${TORCH_LIBRARIES}" "${XLSXWRITER_LIB}" ArffFiles mdlp)
+target_link_libraries(b_best Boost::boost "${XLSXWRITER_LIB}")
+target_link_libraries(b_list ArffFiles mdlp "${TORCH_LIBRARIES}")
+target_link_libraries(testx ArffFiles BayesNet "${TORCH_LIBRARIES}")

src/Platform/Colors.h

@@ -0,0 +1,14 @@
+#ifndef COLORS_H
+#define COLORS_H
+class Colors {
+public:
+    static std::string MAGENTA() { return "\033[1;35m"; }
+    static std::string BLUE() { return "\033[1;34m"; }
+    static std::string CYAN() { return "\033[1;36m"; }
+    static std::string GREEN() { return "\033[1;32m"; }
+    static std::string YELLOW() { return "\033[1;33m"; }
+    static std::string RED() { return "\033[1;31m"; }
+    static std::string WHITE() { return "\033[1;37m"; }
+    static std::string RESET() { return "\033[0m"; }
+};
+#endif // COLORS_H

src/Platform/Dataset.cc

@@ -0,0 +1,215 @@
+#include "Dataset.h"
+#include "ArffFiles.h"
+#include <fstream>
+namespace platform {
+    Dataset::Dataset(const Dataset& dataset) : path(dataset.path), name(dataset.name), className(dataset.className), n_samples(dataset.n_samples), n_features(dataset.n_features), features(dataset.features), states(dataset.states), loaded(dataset.loaded), discretize(dataset.discretize), X(dataset.X), y(dataset.y), Xv(dataset.Xv), Xd(dataset.Xd), yv(dataset.yv), fileType(dataset.fileType)
+    {
+    }
+    string Dataset::getName() const
+    {
+        return name;
+    }
+    string Dataset::getClassName() const
+    {
+        return className;
+    }
+    vector<string> Dataset::getFeatures() const
+    {
+        if (loaded) {
+            return features;
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    int Dataset::getNFeatures() const
+    {
+        if (loaded) {
+            return n_features;
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    int Dataset::getNSamples() const
+    {
+        if (loaded) {
+            return n_samples;
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    map<string, vector<int>> Dataset::getStates() const
+    {
+        if (loaded) {
+            return states;
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    pair<vector<vector<float>>&, vector<int>&> Dataset::getVectors()
+    {
+        if (loaded) {
+            return { Xv, yv };
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    pair<vector<vector<int>>&, vector<int>&> Dataset::getVectorsDiscretized()
+    {
+        if (loaded) {
+            return { Xd, yv };
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    pair<torch::Tensor&, torch::Tensor&> Dataset::getTensors()
+    {
+        if (loaded) {
+            buildTensors();
+            return { X, y };
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    void Dataset::load_csv()
+    {
+        ifstream file(path + "/" + name + ".csv");
+        if (file.is_open()) {
+            string line;
+            getline(file, line);
+            vector<string> tokens = split(line, ',');
+            features = vector<string>(tokens.begin(), tokens.end() - 1);
+            if (className == "-1") {
+                className = tokens.back();
+            }
+            for (auto i = 0; i < features.size(); ++i) {
+                Xv.push_back(vector<float>());
+            }
+            while (getline(file, line)) {
+                tokens = split(line, ',');
+                for (auto i = 0; i < features.size(); ++i) {
+                    Xv[i].push_back(stof(tokens[i]));
+                }
+                yv.push_back(stoi(tokens.back()));
+            }
+            file.close();
+        } else {
+            throw invalid_argument("Unable to open dataset file.");
+        }
+    }
+    void Dataset::computeStates()
+    {
+        for (int i = 0; i < features.size(); ++i) {
+            states[features[i]] = vector<int>(*max_element(Xd[i].begin(), Xd[i].end()) + 1);
+            auto item = states.at(features[i]);
+            iota(begin(item), end(item), 0);
+        }
+        states[className] = vector<int>(*max_element(yv.begin(), yv.end()) + 1);
+        iota(begin(states.at(className)), end(states.at(className)), 0);
+    }
+    void Dataset::load_arff()
+    {
+        auto arff = ArffFiles();
+        arff.load(path + "/" + name + ".arff", className);
+        // Get Dataset X, y
+        Xv = arff.getX();
+        yv = arff.getY();
+        // Get className & Features
+        className = arff.getClassName();
+        auto attributes = arff.getAttributes();
+        transform(attributes.begin(), attributes.end(), back_inserter(features), [](const auto& attribute) { return attribute.first; });
+    }
+    vector<string> tokenize(string line)
+    {
+        vector<string> tokens;
+        for (auto i = 0; i < line.size(); ++i) {
+            if (line[i] == ' ' || line[i] == '\t' || line[i] == '\n') {
+                string token = line.substr(0, i);
+                tokens.push_back(token);
+                line.erase(line.begin(), line.begin() + i + 1);
+                i = 0;
+                while (line[i] == ' ' || line[i] == '\t' || line[i] == '\n')
+                    line.erase(line.begin(), line.begin() + i + 1);
+            }
+        }
+        if (line.size() > 0) {
+            tokens.push_back(line);
+        }
+        return tokens;
+    }
+    void Dataset::load_rdata()
+    {
+        ifstream file(path + "/" + name + "_R.dat");
+        if (file.is_open()) {
+            string line;
+            getline(file, line);
+            line = ArffFiles::trim(line);
+            vector<string> tokens = tokenize(line);
+            transform(tokens.begin(), tokens.end() - 1, back_inserter(features), [](const auto& attribute) { return ArffFiles::trim(attribute); });
+            if (className == "-1") {
+                className = ArffFiles::trim(tokens.back());
+            }
+            for (auto i = 0; i < features.size(); ++i) {
+                Xv.push_back(vector<float>());
+            }
+            while (getline(file, line)) {
+                tokens = tokenize(line);
+                // We have to skip the first token, which is the instance number.
+                for (auto i = 1; i < features.size() + 1; ++i) {
+                    const float value = stof(tokens[i]);
+                    Xv[i - 1].push_back(value);
+                }
+                yv.push_back(stoi(tokens.back()));
+            }
+            file.close();
+        } else {
+            throw invalid_argument("Unable to open dataset file.");
+        }
+    }
+    void Dataset::load()
+    {
+        if (loaded) {
+            return;
+        }
+        if (fileType == CSV) {
+            load_csv();
+        } else if (fileType == ARFF) {
+            load_arff();
+        } else if (fileType == RDATA) {
+            load_rdata();
+        }
+        if (discretize) {
+            Xd = discretizeDataset(Xv, yv);
+            computeStates();
+        }
+        n_samples = Xv[0].size();
+        n_features = Xv.size();
+        loaded = true;
+    }
+    void Dataset::buildTensors()
+    {
+        if (discretize) {
+            X = torch::zeros({ static_cast<int>(n_features), static_cast<int>(n_samples) }, torch::kInt32);
+        } else {
+            X = torch::zeros({ static_cast<int>(n_features), static_cast<int>(n_samples) }, torch::kFloat32);
+        }
+        for (int i = 0; i < features.size(); ++i) {
+            if (discretize) {
+                X.index_put_({ i,  "..." }, torch::tensor(Xd[i], torch::kInt32));
+            } else {
+                X.index_put_({ i,  "..." }, torch::tensor(Xv[i], torch::kFloat32));
+            }
+        }
+        y = torch::tensor(yv, torch::kInt32);
+    }
+    vector<mdlp::labels_t> Dataset::discretizeDataset(vector<mdlp::samples_t>& X, mdlp::labels_t& y)
+    {
+        vector<mdlp::labels_t> Xd;
+        auto fimdlp = mdlp::CPPFImdlp();
+        for (int i = 0; i < X.size(); i++) {
+            fimdlp.fit(X[i], y);
+            mdlp::labels_t& xd = fimdlp.transform(X[i]);
+            Xd.push_back(xd);
+        }
+        return Xd;
+    }
+}

src/Platform/Dataset.h

@@ -0,0 +1,80 @@
+#ifndef DATASET_H
+#define DATASET_H
+#include <torch/torch.h>
+#include <map>
+#include <vector>
+#include <string>
+#include "CPPFImdlp.h"
+#include "Utils.h"
+namespace platform {
+    using namespace std;
+
+    enum fileType_t { CSV, ARFF, RDATA };
+    class SourceData {
+    public:
+        SourceData(string source)
+        {
+            if (source == "Surcov") {
+                path = "datasets/";
+                fileType = CSV;
+            } else if (source == "Arff") {
+                path = "datasets/";
+                fileType = ARFF;
+            } else if (source == "Tanveer") {
+                path = "data/";
+                fileType = RDATA;
+            } else {
+                throw invalid_argument("Unknown source.");
+            }
+        }
+        string getPath()
+        {
+            return path;
+        }
+        fileType_t getFileType()
+        {
+            return fileType;
+        }
+    private:
+        string path;
+        fileType_t fileType;
+    };
+    class Dataset {
+    private:
+        string path;
+        string name;
+        fileType_t fileType;
+        string className;
+        int n_samples{ 0 }, n_features{ 0 };
+        vector<string> features;
+        map<string, vector<int>> states;
+        bool loaded;
+        bool discretize;
+        torch::Tensor X, y;
+        vector<vector<float>> Xv;
+        vector<vector<int>> Xd;
+        vector<int> yv;
+        void buildTensors();
+        void load_csv();
+        void load_arff();
+        void load_rdata();
+        void computeStates();
+        vector<mdlp::labels_t> discretizeDataset(vector<mdlp::samples_t>& X, mdlp::labels_t& y);
+    public:
+        Dataset(const string& path, const string& name, const string& className, bool discretize, fileType_t fileType) : path(path), name(name), className(className), discretize(discretize), loaded(false), fileType(fileType) {};
+        explicit Dataset(const Dataset&);
+        string getName() const;
+        string getClassName() const;
+        vector<string> getFeatures() const;
+        map<string, vector<int>> getStates() const;
+        pair<vector<vector<float>>&, vector<int>&> getVectors();
+        pair<vector<vector<int>>&, vector<int>&> getVectorsDiscretized();
+        pair<torch::Tensor&, torch::Tensor&> getTensors();
+        int getNFeatures() const;
+        int getNSamples() const;
+        void load();
+        const bool inline isLoaded() const { return loaded; };
+    };
+};
+
+#endif

src/Platform/Datasets.cc

@@ -0,0 +1,129 @@
+#include "Datasets.h"
+#include <fstream>
+namespace platform {
+    void Datasets::load()
+    {
+        auto sd = SourceData(sfileType);
+        fileType = sd.getFileType();
+        path = sd.getPath();
+        ifstream catalog(path + "all.txt");
+        if (catalog.is_open()) {
+            string line;
+            while (getline(catalog, line)) {
+                if (line.empty() || line[0] == '#') {
+                    continue;
+                }
+                vector<string> tokens = split(line, ',');
+                string name = tokens[0];
+                string className;
+                if (tokens.size() == 1) {
+                    className = "-1";
+                } else {
+                    className = tokens[1];
+                }
+                datasets[name] = make_unique<Dataset>(path, name, className, discretize, fileType);
+            }
+            catalog.close();
+        } else {
+            throw invalid_argument("Unable to open catalog file. [" + path + "all.txt" + "]");
+        }
+    }
+    vector<string> Datasets::getNames()
+    {
+        vector<string> result;
+        transform(datasets.begin(), datasets.end(), back_inserter(result), [](const auto& d) { return d.first; });
+        return result;
+    }
+    vector<string> Datasets::getFeatures(const string& name) const
+    {
+        if (datasets.at(name)->isLoaded()) {
+            return datasets.at(name)->getFeatures();
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    map<string, vector<int>> Datasets::getStates(const string& name) const
+    {
+        if (datasets.at(name)->isLoaded()) {
+            return datasets.at(name)->getStates();
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    void Datasets::loadDataset(const string& name) const
+    {
+        if (datasets.at(name)->isLoaded()) {
+            return;
+        } else {
+            datasets.at(name)->load();
+        }
+    }
+    string Datasets::getClassName(const string& name) const
+    {
+        if (datasets.at(name)->isLoaded()) {
+            return datasets.at(name)->getClassName();
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    int Datasets::getNSamples(const string& name) const
+    {
+        if (datasets.at(name)->isLoaded()) {
+            return datasets.at(name)->getNSamples();
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    int Datasets::getNClasses(const string& name)
+    {
+        if (datasets.at(name)->isLoaded()) {
+            auto className = datasets.at(name)->getClassName();
+            if (discretize) {
+                auto states = getStates(name);
+                return states.at(className).size();
+            }
+            auto [Xv, yv] = getVectors(name);
+            return *max_element(yv.begin(), yv.end()) + 1;
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    vector<int> Datasets::getClassesCounts(const string& name) const
+    {
+        if (datasets.at(name)->isLoaded()) {
+            auto [Xv, yv] = datasets.at(name)->getVectors();
+            vector<int> counts(*max_element(yv.begin(), yv.end()) + 1);
+            for (auto y : yv) {
+                counts[y]++;
+            }
+            return counts;
+        } else {
+            throw invalid_argument("Dataset not loaded.");
+        }
+    }
+    pair<vector<vector<float>>&, vector<int>&> Datasets::getVectors(const string& name)
+    {
+        if (!datasets[name]->isLoaded()) {
+            datasets[name]->load();
+        }
+        return datasets[name]->getVectors();
+    }
+    pair<vector<vector<int>>&, vector<int>&> Datasets::getVectorsDiscretized(const string& name)
+    {
+        if (!datasets[name]->isLoaded()) {
+            datasets[name]->load();
+        }
+        return datasets[name]->getVectorsDiscretized();
+    }
+    pair<torch::Tensor&, torch::Tensor&> Datasets::getTensors(const string& name)
+    {
+        if (!datasets[name]->isLoaded()) {
+            datasets[name]->load();
+        }
+        return datasets[name]->getTensors();
+    }
+    bool Datasets::isDataset(const string& name) const
+    {
+        return datasets.find(name) != datasets.end();
+    }
+}

src/Platform/Datasets.h

@@ -0,0 +1,31 @@
+#ifndef DATASETS_H
+#define DATASETS_H
+#include "Dataset.h"
+namespace platform {
+    using namespace std;
+    class Datasets {
+    private:
+        string path;
+        fileType_t fileType;
+        string sfileType;
+        map<string, unique_ptr<Dataset>> datasets;
+        bool discretize;
+        void load(); // Loads the list of datasets
+    public:
+        explicit Datasets(bool discretize, string sfileType) : discretize(discretize), sfileType(sfileType) { load(); };
+        vector<string> getNames();
+        vector<string> getFeatures(const string& name) const;
+        int getNSamples(const string& name) const;
+        string getClassName(const string& name) const;
+        int getNClasses(const string& name);
+        vector<int> getClassesCounts(const string& name) const;
+        map<string, vector<int>> getStates(const string& name) const;
+        pair<vector<vector<float>>&, vector<int>&> getVectors(const string& name);
+        pair<vector<vector<int>>&, vector<int>&> getVectorsDiscretized(const string& name);
+        pair<torch::Tensor&, torch::Tensor&> getTensors(const string& name);
+        bool isDataset(const string& name) const;
+        void loadDataset(const string& name) const;
+    };
+};
+
+#endif

src/Platform/DotEnv.h

@@ -0,0 +1,66 @@
+#ifndef DOTENV_H
+#define DOTENV_H
+#include <string>
+#include <map>
+#include <fstream>
+#include <sstream>
+#include <algorithm>
+#include <iostream>
+#include "Utils.h"
+
+//#include "Dataset.h"
+namespace platform {
+    class DotEnv {
+    private:
+        std::map<std::string, std::string> env;
+        std::string trim(const std::string& str)
+        {
+            std::string result = str;
+            result.erase(result.begin(), std::find_if(result.begin(), result.end(), [](int ch) {
+                return !std::isspace(ch);
+                }));
+            result.erase(std::find_if(result.rbegin(), result.rend(), [](int ch) {
+                return !std::isspace(ch);
+                }).base(), result.end());
+            return result;
+        }
+    public:
+        DotEnv()
+        {
+            std::ifstream file(".env");
+            if (!file.is_open()) {
+                std::cerr << "File .env not found" << std::endl;
+                exit(1);
+            }
+            std::string line;
+            while (std::getline(file, line)) {
+                line = trim(line);
+                if (line.empty() || line[0] == '#') {
+                    continue;
+                }
+                std::istringstream iss(line);
+                std::string key, value;
+                if (std::getline(iss, key, '=') && std::getline(iss, value)) {
+                    env[key] = value;
+                }
+            }
+        }
+        std::string get(const std::string& key)
+        {
+            return env.at(key);
+        }
+        std::vector<int> getSeeds()
+        {
+            auto seeds = std::vector<int>();
+            auto seeds_str = env["seeds"];
+            seeds_str = trim(seeds_str);
+            seeds_str = seeds_str.substr(1, seeds_str.size() - 2);
+            auto seeds_str_split = split(seeds_str, ',');
+            transform(seeds_str_split.begin(), seeds_str_split.end(), back_inserter(seeds), [](const std::string& str) {
+                return stoi(str);
+                });
+            return seeds;
+        }
+    };
+}
+#endif

src/Platform/ExcelFile.cc

@@ -0,0 +1,164 @@
+#include "ExcelFile.h"
+
+namespace platform {
+    ExcelFile::ExcelFile()
+    {
+        setDefault();
+    }
+    ExcelFile::ExcelFile(lxw_workbook* workbook) : workbook(workbook)
+    {
+        setDefault();
+    }
+    void ExcelFile::setDefault()
+    {
+        normalSize = 14; //font size for report body
+        row = 0;
+        colorTitle = 0xB1A0C7;
+        colorOdd = 0xDCE6F1;
+        colorEven = 0xFDE9D9;
+    }
+
+    lxw_workbook* ExcelFile::getWorkbook()
+    {
+        return workbook;
+    }
+    void ExcelFile::setProperties(string title)
+    {
+        char line[title.size() + 1];
+        strcpy(line, title.c_str());
+        lxw_doc_properties properties = {
+            .title = line,
+            .subject = (char*)"Machine learning results",
+            .author = (char*)"Ricardo Montañana Gómez",
+            .manager = (char*)"Dr. J. A. Gámez, Dr. J. M. Puerta",
+            .company = (char*)"UCLM",
+            .comments = (char*)"Created with libxlsxwriter and c++",
+        };
+        workbook_set_properties(workbook, &properties);
+    }
+    lxw_format* ExcelFile::efectiveStyle(const string& style)
+    {
+        lxw_format* efectiveStyle = NULL;
+        if (style != "") {
+            string suffix = row % 2 ? "_odd" : "_even";
+            try {
+                efectiveStyle = styles.at(style + suffix);
+            }
+            catch (const out_of_range& oor) {
+                try {
+                    efectiveStyle = styles.at(style);
+                }
+                catch (const out_of_range& oor) {
+                    throw invalid_argument("Style " + style + " not found");
+                }
+            }
+        }
+        return efectiveStyle;
+    }
+    void ExcelFile::writeString(int row, int col, const string& text, const string& style)
+    {
+        worksheet_write_string(worksheet, row, col, text.c_str(), efectiveStyle(style));
+    }
+    void ExcelFile::writeInt(int row, int col, const int number, const string& style)
+    {
+        worksheet_write_number(worksheet, row, col, number, efectiveStyle(style));
+    }
+    void ExcelFile::writeDouble(int row, int col, const double number, const string& style)
+    {
+        worksheet_write_number(worksheet, row, col, number, efectiveStyle(style));
+    }
+    void ExcelFile::addColor(lxw_format* style, bool odd)
+    {
+        uint32_t efectiveColor = odd ? colorEven : colorOdd;
+        format_set_bg_color(style, lxw_color_t(efectiveColor));
+    }
+    void ExcelFile::createStyle(const string& name, lxw_format* style, bool odd)
+    {
+        addColor(style, odd);
+        if (name == "textCentered") {
+            format_set_align(style, LXW_ALIGN_CENTER);
+            format_set_font_size(style, normalSize);
+            format_set_border(style, LXW_BORDER_THIN);
+        } else if (name == "text") {
+            format_set_font_size(style, normalSize);
+            format_set_border(style, LXW_BORDER_THIN);
+        } else if (name == "bodyHeader") {
+            format_set_bold(style);
+            format_set_font_size(style, normalSize);
+            format_set_align(style, LXW_ALIGN_CENTER);
+            format_set_align(style, LXW_ALIGN_VERTICAL_CENTER);
+            format_set_border(style, LXW_BORDER_THIN);
+            format_set_bg_color(style, lxw_color_t(colorTitle));
+        } else if (name == "result") {
+            format_set_font_size(style, normalSize);
+            format_set_border(style, LXW_BORDER_THIN);
+            format_set_num_format(style, "0.0000000");
+        } else if (name == "time") {
+            format_set_font_size(style, normalSize);
+            format_set_border(style, LXW_BORDER_THIN);
+            format_set_num_format(style, "#,##0.000000");
+        } else if (name == "ints") {
+            format_set_font_size(style, normalSize);
+            format_set_num_format(style, "###,##0");
+            format_set_border(style, LXW_BORDER_THIN);
+        } else if (name == "floats") {
+            format_set_border(style, LXW_BORDER_THIN);
+            format_set_font_size(style, normalSize);
+            format_set_num_format(style, "#,##0.00");
+        }
+    }
+
+    void ExcelFile::createFormats()
+    {
+        auto styleNames = { "text", "textCentered", "bodyHeader", "result", "time", "ints", "floats" };
+        lxw_format* style;
+        for (string name : styleNames) {
+            lxw_format* style = workbook_add_format(workbook);
+            style = workbook_add_format(workbook);
+            createStyle(name, style, true);
+            styles[name + "_odd"] = style;
+            style = workbook_add_format(workbook);
+            createStyle(name, style, false);
+            styles[name + "_even"] = style;
+        }
+
+        // Header 1st line
+        lxw_format* headerFirst = workbook_add_format(workbook);
+        format_set_bold(headerFirst);
+        format_set_font_size(headerFirst, 18);
+        format_set_align(headerFirst, LXW_ALIGN_CENTER);
+        format_set_align(headerFirst, LXW_ALIGN_VERTICAL_CENTER);
+        format_set_border(headerFirst, LXW_BORDER_THIN);
+        format_set_bg_color(headerFirst, lxw_color_t(colorTitle));
+
+        // Header rest
+        lxw_format* headerRest = workbook_add_format(workbook);
+        format_set_bold(headerRest);
+        format_set_align(headerRest, LXW_ALIGN_CENTER);
+        format_set_font_size(headerRest, 16);
+        format_set_align(headerRest, LXW_ALIGN_VERTICAL_CENTER);
+        format_set_border(headerRest, LXW_BORDER_THIN);
+        format_set_bg_color(headerRest, lxw_color_t(colorOdd));
+
+        // Header small
+        lxw_format* headerSmall = workbook_add_format(workbook);
+        format_set_bold(headerSmall);
+        format_set_align(headerSmall, LXW_ALIGN_LEFT);
+        format_set_font_size(headerSmall, 12);
+        format_set_border(headerSmall, LXW_BORDER_THIN);
+        format_set_align(headerSmall, LXW_ALIGN_VERTICAL_CENTER);
+        format_set_bg_color(headerSmall, lxw_color_t(colorOdd));
+
+        // Summary style
+        lxw_format* summaryStyle = workbook_add_format(workbook);
+        format_set_bold(summaryStyle);
+        format_set_font_size(summaryStyle, 16);
+        format_set_border(summaryStyle, LXW_BORDER_THIN);
+        format_set_align(summaryStyle, LXW_ALIGN_VERTICAL_CENTER);
+
+        styles["headerFirst"] = headerFirst;
+        styles["headerRest"] = headerRest;
+        styles["headerSmall"] = headerSmall;
+        styles["summaryStyle"] = summaryStyle;
+    }
+}