Python and C++ code refactoring

CHANGELOG.md

@@ -1,5 +1,13 @@
 # Changelog
 
+## v0.5.2
+
+- Python code is now formatted with `black` and `isort`, and has been refactored according to `PEP 8` style guides.
+- Python code partially commented with `google` docstring format.
+- C++ code partially commented with `doxygen` docstring format.
+- Scott's and Normal Reference Rule's `bandwidth` calculation have been reordered and commented.
+- `ArcOperatorSet::update_incoming_arcs_scores` formulas have been reordered and commented.
+
 ## v0.5.1
 
 - Fixes vcpkg bad hashes ([vcpkg/#38974](https://github.com/microsoft/vcpkg/issues/38974)).
@@ -101,7 +109,7 @@ build process is simpler and orchestrated by scikit-build-core and a CMakeLists.
 - Added a `ProductKDE` class that implements `KDE` with diagonal bandwidth matrix.
 - Added an abstract class `BandwidthSelector` to implement bandwidth selection for `KDE` and `ProductKDE`. Three
   concrete implementations of bandwidth selection are included: `ScottsBandwidth`, `NormalReferenceRule` and `UCV`.
-- Added `Arguments`, `Args` and `Kwargs` to store a set of arguments to be used to create new factors through
+- Added `arguments`, `args` and `kwargs` to store a set of arguments to be used to create new factors through
   `FactorType::new_factor()`. The `Arguments` are accepted by `BayesianNetworkBase::fit()` and the constructors of
   `CVLikelihood`, `HoldoutLikelihood` and `ValidatedLikelihood`.
 
@@ -113,8 +121,8 @@ build process is simpler and orchestrated by scikit-build-core and a CMakeLists.
 ## v0.2.0
 
 - Added conditional linear Gaussian networks (`CLGNetworkType`, `CLGNetwork`, `ConditionalCLGNetwork` and `DynamicCLGNetwork`).
-- Implemented `ChiSquare` (and `DynamicChiSquare`) indepencence test.
-- Implemented `MutualInformation` (and `DynamicMutualInformation`) indepencence test. This is valid for hybrid data.
+- Implemented `ChiSquare` (and `DynamicChiSquare`) independence test.
+- Implemented `MutualInformation` (and `DynamicMutualInformation`) independence test. This is valid for hybrid data.
 - Implemented `BDe` (Bayesian Dirichlet equivalent) score (and `DynamicBDe`).
 - Added `UnknownFactorType` as default `FactorType` for Bayesian networks when the node type could not be deduced.
 - Added `Assignment` class to represent the assignment of values to variables.

CMakeLists.txt

@@ -1,67 +1,82 @@
 cmake_minimum_required(VERSION 3.20.0)
 
-IF(WIN32)
+if(WIN32)
     set(VCPKG_TARGET_TRIPLET x64-windows-static)
-ENDIF()
-
-IF(APPLE)
-    SET(CMAKE_C_COMPILER "clang")
-    SET(CMAKE_CXX_COMPILER "clang++")
-ENDIF()
+endif()
+
+if(APPLE)
+    set(CMAKE_C_COMPILER "clang")
+    set(CMAKE_CXX_COMPILER "clang++")
+endif()
+
+if(UNIX)
+    set(CMAKE_C_COMPILER "gcc")
+    set(CMAKE_CXX_COMPILER "g++")
+endif()
+
+# Enable ccache if available
+find_program(CCACHE_PROGRAM ccache)
+if(CCACHE_PROGRAM)
+    message(STATUS "ccache found: ${CCACHE_PROGRAM}")
+    set(CMAKE_C_COMPILER_LAUNCHER "${CCACHE_PROGRAM}")
+    set(CMAKE_CXX_COMPILER_LAUNCHER "${CCACHE_PROGRAM}")
+else()
+    message(STATUS "ccache not found. Compilation will proceed without caching.")
+endif()
 
 find_package(Git REQUIRED)
 message("Git executable: ${GIT_EXECUTABLE}")
 
-IF(EXISTS ".git")
-    SET(GIT_COMMAND_EXECUTED "{GIT_EXECUTABLE} submodule update --init --recursive")
+if(EXISTS ".git")
+    set(GIT_COMMAND_EXECUTED "{GIT_EXECUTABLE} submodule update --init --recursive")
     execute_process(COMMAND ${GIT_EXECUTABLE} submodule update --init --recursive
-                    WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR} RESULT_VARIABLE GIT_SUBMOD_RESULT)
+        WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR} RESULT_VARIABLE GIT_SUBMOD_RESULT)
 
-    IF(NOT GIT_SUBMOD_RESULT EQUAL "0")
+    if(NOT GIT_SUBMOD_RESULT EQUAL "0")
         message(FATAL_ERROR "${GIT_COMMAND_EXECUTED} failed with ${GIT_SUBMOD_RESULT}.")
-    ENDIF()
-ELSE()
-    SET(GIT_COMMAND_EXECUTED "${GIT_EXECUTABLE} clone https://github.com/Microsoft/vcpkg.git")
+    endif()
+else()
+    set(GIT_COMMAND_EXECUTED "${GIT_EXECUTABLE} clone https://github.com/Microsoft/vcpkg.git")
     execute_process(COMMAND ${GIT_EXECUTABLE} clone https://github.com/Microsoft/vcpkg.git
-                    WORKING_DIRECTORY "." RESULT_VARIABLE GIT_SUBMOD_RESULT)
+        WORKING_DIRECTORY "." RESULT_VARIABLE GIT_SUBMOD_RESULT)
 
-    IF(NOT GIT_SUBMOD_RESULT EQUAL "0")
+    if(NOT GIT_SUBMOD_RESULT EQUAL "0")
         message(FATAL_ERROR "${GIT_COMMAND_EXECUTED} failed with ${GIT_SUBMOD_RESULT}.")
-    ENDIF()
-ENDIF()
+    endif()
+endif()
 
-SET(GIT_COMMIT_HASH "2024.08.23")
+set(GIT_COMMIT_HASH "2024.08.23")
 
-SET(GIT_COMMAND_EXECUTED "${GIT_EXECUTABLE} checkout ${GIT_COMMIT_HASH}")
+set(GIT_COMMAND_EXECUTED "${GIT_EXECUTABLE} checkout ${GIT_COMMIT_HASH}")
 execute_process(COMMAND ${GIT_EXECUTABLE} checkout ${GIT_COMMIT_HASH}
-                WORKING_DIRECTORY "vcpkg" RESULT_VARIABLE GIT_SUBMOD_RESULT)
+    WORKING_DIRECTORY "vcpkg" RESULT_VARIABLE GIT_SUBMOD_RESULT)
 
-IF(NOT GIT_SUBMOD_RESULT EQUAL "0")
+if(NOT GIT_SUBMOD_RESULT EQUAL "0")
     message(FATAL_ERROR "${GIT_COMMAND_EXECUTED} failed with ${GIT_SUBMOD_RESULT}.")
-ENDIF()
+endif()
 
-SET(GIT_COMMAND_EXECUTED "${GIT_EXECUTABLE} rev-parse HEAD")
+set(GIT_COMMAND_EXECUTED "${GIT_EXECUTABLE} rev-parse HEAD")
 execute_process(COMMAND ${GIT_EXECUTABLE} rev-parse HEAD
-                                WORKING_DIRECTORY "vcpkg" RESULT_VARIABLE GIT_SUBMOD_RESULT OUTPUT_VARIABLE GIT_STDOUT)
+    WORKING_DIRECTORY "vcpkg" RESULT_VARIABLE GIT_SUBMOD_RESULT OUTPUT_VARIABLE GIT_STDOUT)
 
-IF(NOT GIT_SUBMOD_RESULT EQUAL "0")
+if(NOT GIT_SUBMOD_RESULT EQUAL "0")
     message(FATAL_ERROR "${GIT_COMMAND_EXECUTED} failed with ${GIT_SUBMOD_RESULT}.")
-ENDIF()
+endif()
 
 message("Git commit in vcpkg: ${GIT_STDOUT}")
 
 set(CMAKE_TOOLCHAIN_FILE "vcpkg/scripts/buildsystems/vcpkg.cmake")
 
 project(pybnesian VERSION ${SKBUILD_PROJECT_VERSION} LANGUAGES CXX)
-ADD_DEFINITIONS("-DVERSION_INFO=${SKBUILD_PROJECT_VERSION}")
+add_definitions("-DVERSION_INFO=${SKBUILD_PROJECT_VERSION}")
 
 
 set(CMAKE_CXX_STANDARD 17)
 
-IF(MSVC)
-    SET(CMAKE_MSVC_RUNTIME_LIBRARY "MultiThreaded$<$<CONFIG:Debug>:Debug>")
-    ADD_DEFINITIONS("-DNOGDI")
-ENDIF()
+if(MSVC)
+    set(CMAKE_MSVC_RUNTIME_LIBRARY "MultiThreaded$<$<CONFIG:Debug>:Debug>")
+    add_definitions("-DNOGDI")
+endif()
 
 set(PYBIND11_NEWPYTHON ON)
 find_package(Python COMPONENTS Interpreter Development)
@@ -72,31 +87,35 @@ message("Minor version: ${Python_VERSION_MINOR}")
 
 add_definitions(-DPYTHON_VERSION_MAJOR=${Python_VERSION_MAJOR} -DPYTHON_VERSION_MINOR=${Python_VERSION_MINOR})
 
-IF(WIN32)
-    SET(SCRIPT_PREFIX "")
-    SET(SCRIPT_EXTENSION "bat")
-ELSEIF(UNIX)
-    SET(SCRIPT_PREFIX "./")
-    SET(SCRIPT_EXTENSION "sh")
-ENDIF()
+if(WIN32)
+    set(SCRIPT_PREFIX "")
+    set(SCRIPT_EXTENSION "bat")
+elseif(UNIX)
+    set(SCRIPT_PREFIX "./")
+    set(SCRIPT_EXTENSION "sh")
+endif()
+
+# Find the Python interpreter
+find_package(PythonInterp 3 REQUIRED)
 
-execute_process(COMMAND python expand_sources.py RESULT_VARIABLE EXPAND_SOURCES_RESULT)
+# Use the found Python interpreter in the execute_process command
+execute_process(COMMAND ${PYTHON_EXECUTABLE} expand_sources.py RESULT_VARIABLE EXPAND_SOURCES_RESULT)
 
-IF(NOT EXPAND_SOURCES_RESULT EQUAL "0")
-    message(FATAL_ERROR "$python expand_sources.py failed with ${EXPAND_SOURCES_RESULT}")
-ENDIF()
+if(NOT EXPAND_SOURCES_RESULT EQUAL "0")
+    message(FATAL_ERROR "${PYTHON_EXECUTABLE} expand_sources.py failed with ${EXPAND_SOURCES_RESULT}")
+endif()
 
 execute_process(COMMAND ${SCRIPT_PREFIX}bootstrap-vcpkg.${SCRIPT_EXTENSION} WORKING_DIRECTORY "vcpkg" RESULT_VARIABLE VCPKG_BOOTSTRAP_RESULT)
 
-IF(NOT VCPKG_BOOTSTRAP_RESULT EQUAL "0")
+if(NOT VCPKG_BOOTSTRAP_RESULT EQUAL "0")
     message(FATAL_ERROR "${SCRIPT_PREFIX}bootstrap-vcpkg.${SCRIPT_EXTENSION} failed with ${VCPKG_BOOTSTRAP_RESULT}")
-ENDIF()
+endif()
 
 execute_process(COMMAND ${SCRIPT_PREFIX}vcpkg install WORKING_DIRECTORY "vcpkg" RESULT_VARIABLE VCPKG_INSTALL_RESULT)
 
-IF(NOT VCPKG_INSTALL_RESULT EQUAL "0")
+if(NOT VCPKG_INSTALL_RESULT EQUAL "0")
     message(FATAL_ERROR "${SCRIPT_PREFIX}vcpkg install failed with ${VCPKG_INSTALL_RESULT}")
-ENDIF()
+endif()
 
 find_package(Arrow CONFIG REQUIRED)
 message("Arrow found: ${Arrow_FOUND}")
@@ -117,68 +136,70 @@ find_package(Boost REQUIRED COMPONENTS math dynamic_bitset)
 find_package(OpenCL REQUIRED)
 
 pybind11_add_module(__init__ "pybnesian/lib.cpp"
-                              "pybnesian/pybindings/pybindings_dataset.cpp"
-                              "pybnesian/pybindings/pybindings_kde.cpp"
-                              "pybnesian/pybindings/pybindings_factors.cpp"
-                              "pybnesian/pybindings/pybindings_graph.cpp"
-                              "pybnesian/pybindings/pybindings_models.cpp"
-                              "pybnesian/pybindings/pybindings_learning/pybindings_learning.cpp"
-                              "pybnesian/pybindings/pybindings_learning/pybindings_scores.cpp"
-                              "pybnesian/pybindings/pybindings_learning/pybindings_independences.cpp"
-                              "pybnesian/pybindings/pybindings_learning/pybindings_parameters.cpp"
-                              "pybnesian/pybindings/pybindings_learning/pybindings_mle.cpp"
-                              "pybnesian/pybindings/pybindings_learning/pybindings_operators.cpp"
-                              "pybnesian/pybindings/pybindings_learning/pybindings_algorithms.cpp"
-                              "pybnesian/kde/KDE.cpp"
-                              "pybnesian/kde/ProductKDE.cpp"
-                              "pybnesian/kde/UCV.cpp"
-                              "pybnesian/factors/continuous/LinearGaussianCPD.cpp"
-                              "pybnesian/factors/continuous/CKDE.cpp"
-                              "pybnesian/factors/discrete/DiscreteFactor.cpp"
-                              "pybnesian/factors/discrete/discrete_indices.cpp"
-                              "pybnesian/dataset/dataset.cpp"
-                              "pybnesian/dataset/dynamic_dataset.cpp"
-                              "pybnesian/dataset/crossvalidation_adaptator.cpp"
-                              "pybnesian/dataset/holdout_adaptator.cpp"
-                              "pybnesian/util/arrow_types.cpp"
-                              "pybnesian/util/bit_util.cpp"
-                              "pybnesian/util/validate_options.cpp"
-                              "pybnesian/util/validate_whitelists.cpp"
-                              "pybnesian/util/temporal.cpp"
-                              "pybnesian/util/rpoly.cpp"
-                              "pybnesian/util/vech_ops.cpp"
-                              "pybnesian/util/pickle.cpp"
-                              "pybnesian/util/util_types.cpp"
-                              "pybnesian/kdtree/kdtree.cpp"
-                              "pybnesian/learning/operators/operators.cpp"
-                              "pybnesian/learning/algorithms/hillclimbing.cpp"
-                              "pybnesian/learning/algorithms/pc.cpp"
-                              "pybnesian/learning/algorithms/mmpc.cpp"
-                              "pybnesian/learning/algorithms/mmhc.cpp"
-                              "pybnesian/learning/algorithms/dmmhc.cpp"
-                              "pybnesian/learning/independences/continuous/linearcorrelation.cpp"
-                              "pybnesian/learning/independences/continuous/mutual_information.cpp"
-                              "pybnesian/learning/independences/continuous/RCoT.cpp"
-                              "pybnesian/learning/independences/discrete/chi_square.cpp"
-                              "pybnesian/learning/independences/hybrid/mutual_information.cpp"
-                              "pybnesian/learning/parameters/mle_LinearGaussianCPD.cpp"
-                              "pybnesian/learning/parameters/mle_DiscreteFactor.cpp"
-                              "pybnesian/learning/scores/bic.cpp"
-                              "pybnesian/learning/scores/bge.cpp"
-                              "pybnesian/learning/scores/bde.cpp"
-                              "pybnesian/learning/scores/cv_likelihood.cpp"
-                              "pybnesian/learning/scores/holdout_likelihood.cpp"
-                              "pybnesian/graph/generic_graph.cpp"
-                              "pybnesian/models/BayesianNetwork.cpp"
-                              "pybnesian/models/GaussianNetwork.cpp"
-                              "pybnesian/models/SemiparametricBN.cpp"
-                              "pybnesian/models/KDENetwork.cpp"
-                              "pybnesian/models/DiscreteBN.cpp"
-                              "pybnesian/models/HomogeneousBN.cpp"
-                              "pybnesian/models/HeterogeneousBN.cpp"
-                              "pybnesian/models/CLGNetwork.cpp"
-                              "pybnesian/models/DynamicBayesianNetwork.cpp"
-                              "pybnesian/opencl/opencl_config.cpp")
+    "pybnesian/pybindings/pybindings_dataset.cpp"
+    "pybnesian/pybindings/pybindings_kde.cpp"
+    "pybnesian/pybindings/pybindings_factors.cpp"
+    "pybnesian/pybindings/pybindings_graph.cpp"
+    "pybnesian/pybindings/pybindings_models.cpp"
+    "pybnesian/pybindings/pybindings_learning/pybindings_learning.cpp"
+    "pybnesian/pybindings/pybindings_learning/pybindings_scores.cpp"
+    "pybnesian/pybindings/pybindings_learning/pybindings_independences.cpp"
+    "pybnesian/pybindings/pybindings_learning/pybindings_parameters.cpp"
+    "pybnesian/pybindings/pybindings_learning/pybindings_mle.cpp"
+    "pybnesian/pybindings/pybindings_learning/pybindings_operators.cpp"
+    "pybnesian/pybindings/pybindings_learning/pybindings_algorithms.cpp"
+    "pybnesian/kde/KDE.cpp"
+    "pybnesian/kde/ProductKDE.cpp"
+    "pybnesian/kde/UCV.cpp"
+    "pybnesian/factors/continuous/LinearGaussianCPD.cpp"
+    "pybnesian/factors/continuous/CKDE.cpp"
+    "pybnesian/factors/discrete/DiscreteFactor.cpp"
+    "pybnesian/factors/discrete/discrete_indices.cpp"
+    "pybnesian/dataset/dataset.cpp"
+    "pybnesian/dataset/dynamic_dataset.cpp"
+    "pybnesian/dataset/crossvalidation_adaptator.cpp"
+    "pybnesian/dataset/holdout_adaptator.cpp"
+    "pybnesian/util/arrow_types.cpp"
+    "pybnesian/util/bit_util.cpp"
+    "pybnesian/util/validate_options.cpp"
+    "pybnesian/util/validate_whitelists.cpp"
+    "pybnesian/util/temporal.cpp"
+    "pybnesian/util/rpoly.cpp"
+    "pybnesian/util/vech_ops.cpp"
+    "pybnesian/util/pickle.cpp"
+    "pybnesian/util/util_types.cpp"
+    "pybnesian/kdtree/kdtree.cpp"
+    "pybnesian/vptree/vptree.cpp"
+    "pybnesian/learning/operators/operators.cpp"
+    "pybnesian/learning/algorithms/hillclimbing.cpp"
+    "pybnesian/learning/algorithms/pc.cpp"
+    "pybnesian/learning/algorithms/mmpc.cpp"
+    "pybnesian/learning/algorithms/mmhc.cpp"
+    "pybnesian/learning/algorithms/dmmhc.cpp"
+    "pybnesian/learning/independences/continuous/linearcorrelation.cpp"
+    "pybnesian/learning/independences/continuous/mutual_information.cpp"
+    "pybnesian/learning/independences/continuous/RCoT.cpp"
+    "pybnesian/learning/independences/discrete/chi_square.cpp"
+    "pybnesian/learning/independences/hybrid/mutual_information.cpp"
+    "pybnesian/learning/independences/hybrid/mixed_knncmi.cpp"
+    "pybnesian/learning/parameters/mle_LinearGaussianCPD.cpp"
+    "pybnesian/learning/parameters/mle_DiscreteFactor.cpp"
+    "pybnesian/learning/scores/bic.cpp"
+    "pybnesian/learning/scores/bge.cpp"
+    "pybnesian/learning/scores/bde.cpp"
+    "pybnesian/learning/scores/cv_likelihood.cpp"
+    "pybnesian/learning/scores/holdout_likelihood.cpp"
+    "pybnesian/graph/generic_graph.cpp"
+    "pybnesian/models/BayesianNetwork.cpp"
+    "pybnesian/models/GaussianNetwork.cpp"
+    "pybnesian/models/SemiparametricBN.cpp"
+    "pybnesian/models/KDENetwork.cpp"
+    "pybnesian/models/DiscreteBN.cpp"
+    "pybnesian/models/HomogeneousBN.cpp"
+    "pybnesian/models/HeterogeneousBN.cpp"
+    "pybnesian/models/CLGNetwork.cpp"
+    "pybnesian/models/DynamicBayesianNetwork.cpp"
+    "pybnesian/opencl/opencl_config.cpp")
 
 
 target_include_directories(__init__ PRIVATE "pybnesian")
@@ -187,4 +208,3 @@ target_include_directories(__init__ SYSTEM PRIVATE "lib/eigen-3.3.7" "lib/indica
 target_link_libraries(__init__ PRIVATE Arrow::arrow_static OpenCL::OpenCL NLopt::nlopt libfort::fort Boost::dynamic_bitset Boost::math)
 
 install(TARGETS __init__ LIBRARY DESTINATION ./pybnesian)
-