Windows wrappers for the Arcade Learning Environment/Xitari.

xitari-Windows are bindings to Deepmind´s xitari, published here:

https://github.com/deepmind/xitari.git

This repository currently builds a strongly-typed .NET API for use from C# and/or F#. The API binding is currently being tested and functionality will be added to facilitate the use of Windows-based deep learning frameworks.

Dependencies

• An environmental variable named SWIG_PATH pointing to Simplified Wrapper and Interface Generator executable (.exe) file. This repo has been tested with Swig 3.0.10, available from here

• VS 2017, including "Desktop development with C++" & ".Net desktop development". This repo has been tested with Visual Studio Community 2017 Version 15.3.2 and 15.3.3

TODO

• Finalize the .NET API
• Create a cleaner C++ API as here
• Review if Rcpp could be leveraged (R-Binding)