Linux | Python 3.13+
An acronym for GETMM's (Group's name) Phaast Heuristic Algorithm for Atomic Structure Tuning, is a software developed with the intention of providing a readable and modular library to execute a set of heuristic algorithms for global-minima search
Install a python environment manager or create a virtual environment with python's built-in venv module. Pyenv is recommended for its simplicity, instalation and usage instructions are available in its Repo.
Python version 3.13 and above are needed and it is HIGHLY recommended that a free-threaded python version is used (Example: 3.13.11t in Pyenv), or else the structure comparison procedure total time will be higher than geometry optimization.
GET-PHAAST only needs cython as mandatory dependency so, for the headless usage of phaast, it can be installed in your virtual environment with:
pip install cython
If the GUI is needed, then a full dependency command is installed as:
pip install cython pyopengl pyopengl-accelerate pyopengltk
xTB is needed in your $PATH for basic
searches with the built-in genetic algorithm
Searches using the optional DFT package will also require inserting
Orca in your $PATH
Then the framework is simply installed using pip in the repository, for example:
pip install .
If your workdir is the repository directory.
A PyPI library with pre-compiled binaries will be added a little after the repository is made public.
Examples are present in the examples directory to show how the library can be
used.
The command-line allows a user to run GET-PHAAST's genetic algorithm with just
one obligatory flags (e.g. phaast C6H6).
Additional optional flags are present in the Manual
The GET-PHAAST's GUI, a tool to manage structures as well as genetic algorithm
results, can be accessed through the phaast-gui command.