This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| Psi4NumPy | |
|---|---|
| Name | Psi4NumPy |
| Developer | Psi4, NumPy |
| Programming language | Python (programming language), C++ |
| Operating system | Linux, macOS, Microsoft Windows |
| Genre | Computational chemistry, Scientific computing |
| License | LGPL |
Psi4NumPy is an open-source educational and research-oriented collection of scripts and tutorials that combine the Psi4 quantum chemistry package with the NumPy numerical library to teach and prototype electronic structure methods. It serves as a bridge between high-performance codes like GAMESS (US), NWChem, Gaussian (software), ORCA (chemistry), and approachable teaching materials produced by projects such as Psi4 tutorials and textbooks. The project emphasizes reproducibility and clarity for students, researchers, and developers from institutions like MIT, Princeton University, and University of California, Berkeley.
Psi4NumPy provides didactic implementations of ab initio quantum chemistry algorithms drawn from foundational works by researchers affiliated with John Pople, Walter Kohn, Ronald Dreizler, Martin Head-Gordon, and groups at Lawrence Berkeley National Laboratory. The repository pairs the Psi4 program's integrals and SCF engines with NumPy array operations to enable transparent reproductions of methods such as Hartree–Fock, Møller–Plesset perturbation theory, and coupled-cluster, referencing canonical developments associated with Per-Olov Löwdin, John C. Slater, Erich Hückel, and Pople's group. Tutorials often cite influential works and curricula used at Caltech, Harvard University, and ETH Zurich.
The collection implements core algorithms including restricted and unrestricted self-consistent field procedures inspired by methods developed at University of Cambridge and Columbia University, second-order perturbation techniques tied to concepts from C. A. Coulson, and coupled-cluster formulations reflecting advances from Geert Jan van Oldenborgh and Takeshi Nakajima. Psi4NumPy exposes integral transformation routines, density-fitting techniques related to ideas from Francesco Ragni, and excited-state methods building on results by Anna Krylov and Stefan Hirata. Interoperability features facilitate comparisons with codes like Molpro, Q-Chem, and DFT+U implementations found in community codes maintained at Argonne National Laboratory.
The architecture delegates one-electron and two-electron integral evaluation to the underlying Psi4 C++ libraries while performing tensor contractions and linear algebra via NumPy and BLAS/LAPACK backends common to Intel Corporation optimized distributions. The codebase leverages Python bindings similar to approaches used in PySCF and follows packaging patterns observed in projects hosted by GitHub and curated by communities at Open Science Grid. Design choices reflect reproducibility initiatives promoted by Force11 and Reproducible Research advocates, with modular notebooks and scripts that mirror lecture materials from University of Oxford and University of Toronto courses.
Tutorials walk users through end-to-end examples: geometry specification reminiscent of exercises from IUPAC recommendations, SCF convergence studies found in coursework at Imperial College London, correlation energy calculations paralleling exercises used at Stanford University, and visualization workflows integrating tools from Jupyter Project, Matplotlib, and VMD (software). Sample notebooks illustrate stepwise construction of molecular Hamiltonians, integral transformations, and perturbative corrections, drawing pedagogical lineage from texts by Fukui, Szabo and Ostlund, and lecture notes used by David Sherrill and Trygve Helgaker.
Development is collaborative, with contributions from academic groups at Yale University, University of Illinois Urbana-Champaign, University of Wisconsin–Madison, and national labs such as Brookhaven National Laboratory. The community coordinates via platforms and events like GitHub, Zenodo, and conferences including American Chemical Society meetings and the Gordon Research Conferences. Educational adopters include summer schools affiliated with PRACE and workshop series organized by CITATION NEEDED-style academic consortia; maintainers provide issue tracking, continuous integration, and example galleries consistent with best practices endorsed by NumFOCUS.
Psi4NumPy is distributed under permissive free-software terms compatible with scientific redistribution and integration with software from entities like Open Source Initiative, following license models akin to those used by GNU Project derivatives and other academic packages. Binary and source distribution channels include package managers referenced by communities at Conda (package manager), PyPI, and container images used in reproducible deployment at Docker, Inc.. The licensing approach supports academic, industrial, and governmental use cases encountered in collaborations with organizations such as European Commission funded projects and national research programs.
Category:Computational chemistry software