ymahlau/fdtdx

AI Summary: Optimize the FDTDX Python package for speed by implementing a DiamondCandy-based parallelization scheme. This involves restructuring the code to handle smaller computational units, adapting source/detector/boundary handling, and ensuring compatibility with JAX's automatic differentiation for gradient computation. The goal is to significantly reduce simulation and gradient computation time, especially on multi-GPU systems.

AI Summary: Implement a `same_position` function in the FDTDX Python package to simplify the process of placing objects at identical coordinates, improving user experience.

AI Summary: The task is to remove the `fdtdx.SimulationState` class from the FDTDX Python package. This class is deemed unnecessary complexity as it's a simple wrapper for an integer and arrays. The removal should be done carefully to avoid breaking existing code that uses this class.

AI Summary: Enhance the FDTDX Python package's logging functionality to include the user's script in the generated 'code.zip' archive. This will improve reproducibility by archiving the complete simulation setup. Consider either automatically including the main script or allowing users to specify additional files for inclusion.