A python computational physics framework.
Minimal version of pylada necessary to just run the crystal,VASP,ewald,jobs,and database modules
Constructed by Peter Graf from Mayeul d'Avezac's pylada
There are currently two pre-requisites:
Both are generally available on Linux and OS/X (via homebrew for instance).
The simplest approach is to install via pip:
global installation
pip install git+https://github.com/pylada/pylada-light
local (user) installation
pip install --user git+https://github.com/pylada/pylada-light
in a virtual environment
mkvirtualenv --system-site-packages pylada source pylada/bin/activate pip install git+https://github.com/pylada/pylada-light
This last approach is recommended since it keeps the pylada environment isolated from the rest of the system. Susbsequently, this environment can be accessed by running the second line.
There are two approaches for developping with pylada. One is to use the bare-bone cmake build system. The other is to two have pip setup the cmake system for you.
In either case, the source should first be obtained from the github repo.
bare-bone
cd /path/to/source mkdir build cd build cmake .. make make test
The usual cmake options apply. In order to facilitate debugging, a script :bash:`localpython.sh` exists in the build directory to a run a python session that will know where to find the pylada package that is being developped. For instance :bash:`./localpython.sh -m IPython` will launch an ipython session where the current pylada can be imported.
pip develop
mkvirtualenv --system-site-packages pylada source pylada/bin/activate pip install -e git+https://github.com/pylada/pylada-light#egg=pylada cd pylada/src/pylada/build make test
The above creates a virtual environment and installs pylada inside it in development mode. This means that the virtual environment will know about the pylada flavor in development. It is possible to edit a file, do :bash:`make`, launch python and debug. One just needs to active the virtual environment once per session.