PySEAMS: Pythonic Structural Elucidation Analysis for Molecular Simulations

Computational representations of particle data over time or configuration space form a lion's share of high performance computing. For a wealth of ranges, from ab-inito simulations to molecular dynamics for phase transitions and even spectroscopic data, efficient (near-realtime) post-processing requirements lead to a "dual-programming" paradigm, wherein the algorithms are implemented in a statically compiled, performant programming language, while for flexibility and ease of use, the user interface is in a dynamic, scripting language. The predecessor of PySEAMS, d-SEAMS had an embedded `lua` interpreter for facilitating user interactions. The PySEAMS project, by rewriting the internal binding logic to allow first class generation of the base structures used in the subsequent ring-determination analysis, represents a broadening of its applications and allows for more flexible workflows including graph neural networks and more statistical learning techniques.