Bindings: How C++ Compiling Works for Python and R

C++ Core Development

You write and edit your C++ code (e.g., in src/).

Compiling for Bindings

• To use your C++ code in Python or R, you must compile it into a shared library (e.g., .so on Linux, .dll on Windows, .dylib on macOS).
• This is usually done with CMake or a similar build system, which creates the shared library and the necessary binding code (using pybind11 for Python, Rcpp for R).

Python Package

• The Python package (created with pybind11) includes the compiled shared library (e.g., gridfit.cpython-310-x86_64-linux-gnu.so).
• When you install the Python package (e.g., with pip install .), the build process compiles the C++ code and places the shared library in the package directory.
• Python imports this shared library as a module.

R Package

• The R package (created with Rcpp) also includes the compiled shared library (e.g., gridfit.so).
• When you build/install the R package (e.g., with devtools::install()), R compiles the C++ code and places the shared library in the package’s libs/ directory.
• R loads this shared library when you load the package.

Summary

• The C++ code is compiled into a shared library for both Python and R.
• The packages (Python and R) include the compiled version, not the raw C++ source (except for development or source distributions).
• You need to recompile whenever you change the C++ code and want those changes available in Python or R.