Python
- Description
- Python is a cross platform dynamic object-oriented programming language.
- Modulefile
python- Documentation
- Python homepage
- Usage
$ module load python$ module load gcc python/2.7.2$ module load gcc python/3.4.1- You will need also to load the
openmpi/gccmodulefile if you are going to use mpi4py.
Contents
Python Wheelhouse on Glacdyn/Glacdyn2
Similar to the Compute Canada Software Stack we have started a small Wheelhouse with Python wheels on the legacy ACENET machines. It is located at /usr/local/wheelhouse and can be activated by setting the PIP_CONFIG_FILE variable as shown below.
The following example creates a virtual python environment and installs a Cartopy wheel, that has been tested on our systems:
# load PIP configuration to activate wheelhouse: export PIP_CONFIG_FILE=/usr/local/pip.conf # load required modules: module purge module load gcc/4.9.3 python/3.7.7 proj/8.2.1 geos/3.10.2 # create and activate venv python3 -m venv ~/env_cartopy source ~/env_cartopy/bin/activate # install packages from wheelhouse (--no-index bypasses PyPI) pip install --no-index -U pip wheel numpy scipy matplotlib ipython Cartopy==0.20.2
Python modules
SciPy Stack
| Python module | 2.6.2 | 2.7.2 | 2.7.10 | 3.4.x |
|---|---|---|---|---|
| Numpy | 1.4.1 | 1.8.1 | 1.9.2 | 1.12.0 |
| Scipy | 0.7.2 | 0.14.0 | 0.16.0 | 0.18.1 |
| Matplotlib | 1.1.0 | 1.2.1 | 1.4.3 | 1.5.1 |
| IPython | 0.12 | 1.2.1 | 4.2.0 | 4.2.0 |
| pandas | -- | 0.14.0 | 0.18.1 | 0.18.1 |
| SymPy | -- | 0.7.5 | -- | 1.0 |
| Nose | 0.11.3 | 1.3.3 | 1.3.7 | 1.3.7 |
Other
| Python module | 2.6.2 | 2.7.2 | 2.7.10 | 3.4.1 |
|---|---|---|---|---|
| Mercurial | 2.11 | 2.11 | -- | -- |
| NumericalPython | 23.8.2 | 23.8.2 | -- | -- |
| ScientificPython | 2.8 | 2.8 | -- | -- |
| Stsci_python | 2.9 | -- | -- | -- |
| mpi4py | 1.2.1 | 1.3 | -- | -- |
| numexpr | -- | 2.1 | -- | -- |
| netCDF4 | -- | 1.1.0 | -- | 1.2.4 |
| PyTables | -- | 3.0.0 | -- | -- |
| basemap | -- | 1.0.6 | 1.0.7 | -- |
| SCons | -- | 2.3.4 | -- | -- |
| SIP | -- | 4.19.2 | 4.19.2 | -- |
| PyQt4 | -- | 4.12 | 4.12 | -- |
If you need additional modules, we encourage you to installing them in your own filespace using one of the methods described below:
How to install own modules for Python 3 or Python 2.7
You can use the pip utility for easy installation. See the following page for details (do not forget the --user option), like so:
$ pip install --user package-name
You can use easy_install, too.
How to install own modules for any version of Python
See the Python documentation for details:
- Configure Python to find the new modules by adding the following line into your
~/.bashrc(replacepython2.Xwithpython2.6orpython2.7depending on the version of python you are using):
$ export PYTHONPATH=$HOME/lib/python2.X/site-packages/:$PYTHONPATH
Then you can re-source your ~/.bashrc to get the changes into the current shell avoiding logging out:
$ source ~/.bashrc
- Download and unpack the source files. (Here as an example we download the
xlrdpackage.)
$ wget --no-check-certificate https://pypi.python.org/packages/source/x/xlrd/xlrd-0.9.3.tar.gz $ tar xzf xlrd-0.9.3.tar.gz $ cd xlrd-0.9.3/
- Install a module specifying a prefix
$ python setup.py install --prefix=$HOME
- If a module requires a fortran compiler then we strongly recommend to specify
gfortraninstead ofg77like so:
$ python setup.py build --fcompiler=gnu95 install --prefix=$HOME