/PyTorchDIA

PyTorchDIA is a Difference Image Analysis tool. It is built around the PyTorch machine learning framework, and uses automatic differentiation and (optional) GPU support to perform lightning fast optimisations of image models.

Primary LanguageJupyter NotebookMIT LicenseMIT

PyTorchDIA is an open source, Pythonic, GPU-accelerated numerical DIA algorithm.

For a comprehensive overview, check out the paper https://arxiv.org/abs/2104.13715

Please see the tutorial (a work in progress!) for examples of how to use the code.

And please do feel free to get in touch for any questions, suggestions for improvements etc. Email: jah36@st-andrews.ac.uk

TODO: test on CPU