shogun

Why we included this project

SHOGUN has been around since 1999, and that longevity shows in how complete the toolbox feels. It is built on a single C++ core that generates the same interface across Python, R, Java/Scala, Ruby, C#, Octave, and Lua, so you can train a model in one language and call it from another without maintaining parallel implementations. That matters if you work with SVMs, kernel methods, or the broader family of classical classifiers and regressors and need one consistent algorithm across a multilingual codebase. Because the heavy work runs in compiled C++, it is also a good fit when speed matters or when you want to embed ML logic directly into a C++ application. The project is maintained under the NumFOCUS umbrella, which gives it institutional backing that single-maintainer tools rarely have.