The freud Python library provides a simple, flexible, powerful set of tools for analyzing trajectories obtained from molecular dynamics or Monte Carlo simulations. High performance, parallelized C++ is used to compute standard tools such as radial distribution functions, correlation functions, order parameters, and clusters, as well as original analysis methods including potentials of mean force and torque (PMFTs) and local environment matching. The freud library supports many input formats and outputs NumPy arrays, enabling integration with the scientific Python ecosystem for many typical materials science workflows.
The easiest ways to install freud are using pip:
pip install freud-analysis
conda install -c conda-forge freud
freud is also available via containers for Docker or Singularity. If you need more detailed information or wish to install freud from source, please refer to the Installation Guide to compile freud from source.
Some other helpful links for working with freud:
The freud library is called using Python scripts. Many core features are demonstrated in the freud documentation. The examples come in the form of Jupyter notebooks, which can also be downloaded from the freud examples repository or launched interactively on Binder. Below is a sample script that computes the radial distribution function for a simulation run with HOOMD-blue and saved into a GSD file.
import freud import gsd.hoomd # Create a freud compute object (RDF is the canonical example) rdf = freud.density.RDF(bins=50, r_max=5) # Load a GSD trajectory (see docs for other formats) traj = gsd.hoomd.open('trajectory.gsd', 'rb') for frame in traj: rdf.compute(system=frame, reset=False) # Get bin centers, RDF data from attributes r = rdf.bin_centers y = rdf.rdf
Support and Contribution¶
Please visit our repository on GitHub for the library source code. Any issues or bugs may be reported at our issue tracker, while questions and discussion can be directed to our forum. All contributions to freud are welcomed via pull requests!