GASP

Official Resources

• Homepage: http://gasp.mse.cornell.edu/
• Source Repository: https://github.com/henniggroup/GASP-python
• Documentation: https://github.com/henniggroup/GASP-python/blob/master/manual/manual.pdf
• License: GPL-3.0

Overview

GASP (Genetic Algorithm for Structure and Phase Prediction) is a Python-based evolutionary algorithm package developed by the Hennig Group (Cornell/University of Florida). It is designed to predict stable crystal structures and phase diagrams by interfacing with ab initio (VASP) or classical (LAMMPS, GULP) energy calculators.

Scientific domain: Crystal structure prediction, phase diagram determination, evolutionary algorithms
Target user community: Materials scientists, physicists, method developers

Theoretical Methods

• Genetic Algorithm (GA): Global optimization strategy.
• Grand Canonical GA: Allows variable composition searches for phase diagram prediction.
• Evolutionary Operators: Crossover, mutation, permutation, strain.
• Energy Evaluation: External interfaces to DFT (VASP) or classical potentials.

Capabilities

• Crystal Structure Prediction: Finds low-energy structures for fixed compositions.
• Phase Diagram Prediction: Variable composition search to identify stable stoichiometries (convex hull).
• Potentials Testing: Can be used to fit or test empirical potentials against DFT data.
• Interfacing: Supports VASP, LAMMPS, GULP, and MOPAC.
• Symmmetrization: Can enforce or detect symmetry in generated structures.

Inputs & Outputs

• Input:
  • gasp_input.xml or similar configuration file.
  • Calculator input files (e.g., INCAR, POTCAR for VASP).
• Output:
  • run_data directory containing structure files (POSCAR format).
  • statistics files tracking energy and evolution.
  • best_structures list.

Interfaces & Ecosystem

• VASP: Primary first-principles engine.
• LAMMPS/GULP: For classical forcefield calculations.
• Python: Written in Python 2.7 (legacy) / Python 3 (modern branches).

Verification & Sources

• Confidence: ✅ VERIFIED
• Primary Source: GASP GitHub Repository
• Reference: W. W. Tipton and R. G. Hennig, "A grand canonical genetic algorithm for the prediction of multi-component phase diagrams and testing of empirical potentials", J. Phys.: Condens. Matter 25, 495401 (2013).