Pymatgen (python materials genomics) is the python library that powers the Materials Project (http://www.materialsproject.org). These are some of the main features:
- Highly flexible classes for the representation of Element, Site, Molecule, Structure objects.
- Extensive io capabilities to manipulate many VASP input and output files (http://cms.mpi.univie.ac.at/vasp/) and the crystallographic information file format. This includes generating Structure objects from vasp input and output. There is also support for Gaussian input files and XYZ file for molecules.
- Comprehensive tool to generate and view compositional and grand canonical phase diagrams.
- Electronic structure analyses (DOS and Bandstructure).
The public version of pymatgen is free (as in free beer) to download and to use. However, we would also like you to help us improve this library by making your own contributions as well. These contributions can be in the form of additional tools or modules you develop, or even simple things such as bug reports. Please contact the maintainer of this library (firstname.lastname@example.org) to find out how to include your contributions via github or for bug reports.
Note that pymatgen, like all scientific research, will always be a work in progress. While the development team will always strive to avoid backward incompatible changes, they are sometimes unavoidable, and tough decisions have to be made for the long term health of the code.
For documentation and usage guide, please refer to the latest documentation at our github page (http://materialsproject.github.com/pymatgen/). If you wish to be notified via email of pymatgen releases, you may become a member of pymatgen's Google Groups page (https://groups.google.com/forum/?fromgroups#!forum/pymatgen/).
Required for proper functioning of the code
- Python 2.7+ required. New default modules such as json are used, as well as new unittest features in Python 2.7.
- numpy - For array, matrix and other numerical manipulations. Used extensively by all core modules.
- pyspglib 1.2+ (highly recommended): For symmetry finding. Needed if you are using the pymatgen.symmetry, pymatgen.transformation and pymatgen.alchemy packages. From pymatgen v2.1 onwards, pyspglib should be automatically compiled as an extension during the install process via setup.py.
Optional Python Libraries
Optional python libraries that are required if you need certain features
- scipy 0.10+ (highly recommended) - For use in Gaussian smearing and phase diagram construction using scipy.spatial.Delaunay.
- matplotlib (highly recommended): For plotting (e.g., Phase Diagrams).
- PyCifRW (highly recommended): For reading and writing Crystallographic Information Format (CIF) files. Get it from http://pycifrw.berlios.de/ or a working version is provided in the dependencies directory of pymatgen.
- VTK with Python bindings (http://www.vtk.org/): For visualization of crystal structures using the pymatgen.vis package.
- Atomistic Simulation Environment or ASE : Required for the usage of the adapters in pymatgen.io.aseio between pymatgen's core Structure object and the Atoms object used by ASE. Get it at https://wiki.fysik.dtu.dk/ase/.
- OpenBabel with Python bindings (http://openbabel.org). Required for the usage of the adapters in pymatgen.io.babelio between pymatgen's Molecule and OpenBabel's OBMol. Opens up input and output support for the very large number of input and output formats supported by OpenBabel.
- nose - For complete unittesting. This is NOT optional for developers!
Optional non-Python programs
Optional non-python libraries (because no good pythonic alternative exists at the moment) required only for certain features.
- Qhull : Needed for bond length analysis (structure_analyzer.py), or if you use the use_external_qhull option in phase diagram generation.. The executable qconvex and qvoronoi must be in the path.
- ffmpeg : Needed for generation of movies (structure_vtk.py). The executable ffmpeg must be in the path.
- enum : Needed for the use of EnumerateStructureTransformation and the pymatgen.command_line.enumlib_caller module. This library by Gus Hart provides a robust way to enumerate derivative structures. It can be used to completely enumerate all symmetrically distinct ordered structures of disordered structures via the EnumerateStructureTransformation. The multienum.x and makestr.x executables must be in the path.
Basic Setup for Non-developers
If you are using pymatgen purely as a library and do not intend to contribute code, you may install pymatgen either easy_install or python setup.py.
If you have easy_install in your Python setup, the simplest way to get the latest stable release of pymatgen is to do:
If you don't have easy_install, or you prefer to install the latest development version of pymatgen, you can download the tarball and then do the following:
tar -zxvf pymatgen.tar.gz cd pymatgen python setup.py install
You may need to run the above with root privileges on your machine. In addition, you may need to install additional python libraries and dependencies.
With these two basic steps, you should be able to use most of the pymatgen code.
I recommend that you start by reading some of the unittests in the tests subdirectory for each package. The unittests demonstrate the expected behavior and functionality of the code.
However, some extra functionality do require additional setup, as outlined below.
Setup for Developers
There are two categories of developers. General developers should follow the procedures outlined in the pymatgen documentation on collaborative Github workflow to fork a copy of pymatgen to their own Github accounts and cloning it into their local machine.
Core developers who have write access to the main Github repo may clone the pymatgen repo directly.
For both kinds of developers, it is recommended that after you clone the repo, you either add the pymatgen repo to your PYTHONPATH or use
python setup.py develop
which will install pymatgen in development mode and install some of the necessary dependencies.
For the code to generate POTCAR files, it needs to know where the VASP pseudopotential files are. We are not allowed to distribute these under the VASP license. The good news is that we have included a setup script to help you along.
If you cloned the repo directly from github, you should have a run_me_first.sh file in the root directory of your local repo. Otherwise, you can get it directly from our github site at http://github.com/materialsproject/pymatgen. Run the shell script and follow the instructions. If you have done it correctly, you should get a resources directory with the following directory structure:
- psp_resources |- POT_GGA_PAW_PBE ||- POTCAR.Ac_s.gz ||- POTCAR.Ac.gz ||- POTCAR.Ag.gz ... |- POT_GGA_PAW_PW91 ...
After generating the resources directory, you should add a VASP_PSP_DIR environment variable pointing to the generated directory and you should then be able to generate POTCARs.
Alternatively, you can setup the above directly structure manually and set the VASP_PSP_DIR environment variable accordingly.