Fanpy: A python library for prototyping multideterminant methods in ab initio quantum chemistry

Area of Science:

• Computational chemistry
• Electronic structure theory
• Quantum chemistry

Background:

• Developing and testing multideterminant wavefunctions is crucial for accurate electronic structure calculations.
• Existing frameworks can be complex, hindering rapid prototyping of new methods.

Purpose of the Study:

• Introduce Fanpy, a free and open-source Python library for developing and testing multideterminant wavefunctions.
• Facilitate the rapid prototyping of new ab initio methods in electronic structure theory.
• Provide a modular and accessible platform for computational chemists.

Main Methods:

• Fanpy is based on the Flexible Ansatz for N-electron Configuration Interaction (FANCI) framework.
• Multideterminant wavefunctions are represented by overlaps with Slater determinants of orthonormal spin-orbitals.
• The library features a modular design for wavefunction models, Hamiltonians, and objective functions.

Main Results:

• Fanpy allows users to implement new wavefunction ansätze by writing a function for overlap evaluation.
• Its modularity enables easy mixing of methods and exploration of new ideas.
• The library is written in Python with standard dependencies, ensuring cross-platform accessibility.

Conclusions:

• Fanpy offers a user-friendly and efficient tool for advancing electronic structure theory.
• Its design promotes innovation and accessibility in computational chemistry research.
• The library adheres to modern software development principles for reliability and maintainability.