AutomaticDifferentiation | Often one does not need the full complexity | Math library

AutomaticDifferentiation is a C++ library typically used in Utilities, Math, Pytorch applications. AutomaticDifferentiation has no bugs, it has no vulnerabilities and it has low support. You can download it from GitHub.

Often one does not need the full complexity of an automatic differentiation library, which may be difficult to "dominate" (especially in terms of error messages). Besides that the evolution of C++ makes it increasingly easy to implement your own expression templates and AD. So here I'll write a step-by-step "tutorial" on how to do it in modern C++. This example comes as a follow up from an example of generic programming, which we showed during a C++ course at CSCS a couple of years ago ( towards the end, sorry for the poor quality). In that occasion the goal was to show that using C++14 and constexpr we could write expression templates and automatic differentiation with a little effort and an effective syntax. The goal now is to show how this idiom can evolve using C++17 constexpr lambdas, further reducing the coding effort. DISCLAIMER: the purpose of this repo is to present a proof of concept, so all protections are skipped.