Python bindings for KeOps

We now fully document the public interface of the pykeops module, which is a NumPy and PyTorch front-end for the KeOps C++/CUDA library. This package contains three sets of instructions:

1. The LazyTensor wrapper: our high-level interface, which provides a pythonic support for the most useful features of KeOps.

2. The Genred module: our low-level syntax, which interacts directly with the KeOps++ engine.

3. The generic_reduction functions: a legacy collection of helper routines for the Genred module.

• Python install
  • Requirements
  • Using pip (recommended)
  • On Google Colab
  • Using Docker or Singularity
  • From source using git
  • Testing your installation
  • Troubleshooting
• LazyTensors
  • Overview
  • Documentation
• Generic reductions
  • Overview
  • Documentation
  • An example
• Generic solver
  • Overview
  • Documentation
  • An example
• Math-friendly syntax
  • Documentation
  • An example
• Block-sparse reductions
  • Complexity of the KeOps routines
  • Documentation
  • Examples