1.1. Overview

This page describes high-level architecture and the goals of the project.

1.1.1. Goals

This project is Measure 2 “Open Interfaces” of the Mathematical Research Data Initiative (MaRDI) project. Its goal is to provide interoperability between different languages for scientific computing, such that users of one language could use algorithm implementations from another language.

For example, a user can be mostly comfortable with Python, but desires to use an implementation of a solver for partial differential equations written in C. If the solver doesn’t offer Python bindings, the user needs to write them with tools like ctypes, Python C API, Cython, etc.

With the help of open interfaces, this responsibility can be taken from the user, if this particular solver is supported via Open Interfaces.

Then the necessary data conversions between Python and C are provided automatically, and the time to running the solver from Python is much less.

Moreover, when some particular solver is supported in Open Interfaces, then the bindings are provided in several languages such as Python, Julia, C, C++.

Because the heart of the Open Interfaces library is in the library that provides automatic conversion between different languages, it allows to avoid a situation where for each particular algorithm implementation the authors need to provide bindings for several scientific computing languages, which can be very time-consuming.