JABACAT-created machine learning library from scratch.
Important
This project uses The Meson Build system. Follow instructions posted on the website to install meson. On Windows, there is a winget package available.
To build, setup the build directory. Any configuration steps should be done now (specifying compiler or linker, etc.).
meson setup build # replace `build` with any arbitrary output build directoryThen, compile it.
meson compile -C build # replace `build` with the directory specified in the previous stepThe outputs are in build/ (or whatever directory specified). Install the outputs:
meson install -C build # again, replace `build`This project uses Catch2 as its testing framework.
Run meson test to test the project.
meson test -C build # replace `build`Unfortunately, meson doesn't provide good output when a test fails. Running the test executable manually is often more helpful. Look under build/tests/ to find the right executable. The executable for testing the core library is core_tests.
We use clang-format to check our formatting. Before you push your code, you can run the following that finds all source code files, and then formats them all in place.
find . -iname "*.hpp" -o -iname "*.cpp" -o -iname "*.c" -o -iname "*.h" -o -iname "*.tpp" | xargs -I {} clang-format {}
You can also add this to your editors formatting system or some precommit step.