If possible, we recommend to be explicit about which version to use by specifying either python2 and python3 in your scripts.
Python 2 and Python 3 are both available via python2 and python3. The command python is an alias for python2. We recommend to be explicit about which version you want to use, also when using the default Python 2, i.e. use python2 when you know your script requires Python 2 and python3 when you know it requires Python 3.
The below examples uses Python 3, but it works analogously in Python 2, i.e. just replace python3 with python2.
The standard way to install Python packages is by using the pip package management system. You often find installation instructions online such as:
$ pip install pandas
It will not work. If you attempt to run this as-is on the cluster, you get lots of errors complaining about lack of write permissions etc., which is because it tries to install the package in the system-wide Python package folder (to which only sysadms have write permission). You might also see instructions saying you should use sudo ... - that will also not work for the same reason.
There are two ways for non-privileged users to install Python packages using the ‘pip’ module:
Install globally to your home directory (typically ~/.local/lib/python3.6/site-packages/) using python3 -m pip install --user ...
Install locally to a project-specific folder (e.g. ~/my_project/) using python3 -m pip install ... in a self-contained Python virtual environment
Both are done from the terminal. Which one you choose depends on your needs; for some projects you may want to use the virtual environment approach whereas for your everyday work you might want to work toward your global Python package stack. Installing globally is the easiest, because you don’t have to remember to activate a virtual environment and if you need the Python package in different places, you only have to install it once. However, if you are concerned about reproducibility, or being able to coming back to an old project of yours, you most likely want to use a virtual environment for that project so that its Python packages are not updated when you update or install Python packages globally. This is also true if you collaborate with others in a shared project folder.
First of all, if an online installation instructions says pip install ..., replace that with python3 -m pip install .... Second, to install globally to your home directory, remember to always specify the --user option. For example,
[alice@c4-dev1 ~]$ python3 -m pip install --user pandas
Collecting pandas
Downloading https://files.pythonhosted.org/packages/c3/e2/00cacecafbab071c787019f00ad84ca3185952f6bb9bca9550ed83870d4d/pandas-1.1.5-cp36-cp36m-manylinux1_x86_64.whl (9.5MB)
Collecting python-dateutil>=2.7.3 (from pandas)
Downloading https://files.pythonhosted.org/packages/36/7a/87837f39d0296e723bb9b62bbb257d0355c7f6128853c78955f57342a56d/python_dateutil-2.8.2-py2.py3-none-any.whl (247kB)
Collecting numpy>=1.15.4 (from pandas)
Downloading https://files.pythonhosted.org/packages/45/b2/6c7545bb7a38754d63048c7696804a0d947328125d81bf12beaa692c3ae3/numpy-1.19.5-cp36-cp36m-manylinux1_x86_64.whl (13.4MB)
Collecting pytz>=2017.2 (from pandas)
Downloading https://files.pythonhosted.org/packages/85/ac/92f998fc52a70afd7f6b788142632afb27cd60c8c782d1452b7466603332/pytz-2022.6-py2.py3-none-any.whl (498kB)
Requirement already satisfied: six>=1.5 in /usr/lib/python3.6/site-packages (from python-dateutil>=2.7.3->pandas)
Installing collected packages: python-dateutil, numpy, pytz, pandas
Successfully installed numpy-1.19.5 pandas-1.1.5 python-dateutil-2.8.2 pytz-2022.6
To see all Python packages that you have installed globally, use python3 -m pip list --user. To also see packages installed site wide on the cluster, use python3 -m pip list. Packages installed with python3 -m pip list --user are typically installed to your ~/.local/lib/python3.6/site-packages/ folder. If CLI executables are installed with one of those packages, they are often installed to ~/.local/bin/.
Virtual environment are not used just on computer clusters - many Python users and developers choose to use virtual environment on their local computers whenever they work in Python.
An alternative to install globally to your home directory, is to install to a local folder using a, so called, Python virtual environment. A virtual environment is a self-contained folder that contains the Python executable and any Python packages you install. When you activate a virtual environment, environment variables like PATH is updated such that you will use the Python executable and the packages in the virtual environment and not the globally installed ones.
Below is an example on how to set up a virtual environment and install the pandas package and all of its dependencies into it.
In order to use virtual environments, we need the virtualenv tool. Following the above instructions, you can install it to your global stack as:
[alice@c4-dev1 ~]$ python3 -m pip install --user virtualenv
Collecting virtualenv
Downloading https://files.pythonhosted.org/packages/fd/76/c99d37939e17e3c53f1c1b7e4f5365a9160b1cd0b37700657eed9a2a6775/virtualenv-20.16.7-py3-none-any.whl (8.8MB)
Collecting platformdirs<3,>=2.4 (from virtualenv)
Downloading https://files.pythonhosted.org/packages/b1/78/dcfd84d3aabd46a9c77260fb47ea5d244806e4daef83aa6fe5d83adb182c/platformdirs-2.4.0-py3-none-any.whl
Collecting distlib<1,>=0.3.6 (from virtualenv)
Downloading https://files.pythonhosted.org/packages/76/cb/6bbd2b10170ed991cf64e8c8b85e01f2fb38f95d1bc77617569e0b0b26ac/distlib-0.3.6-py2.py3-none-any.whl (468kB)
Collecting filelock<4,>=3.4.1 (from virtualenv)
Downloading https://files.pythonhosted.org/packages/84/ce/8916d10ef537f3f3b046843255f9799504aa41862bfa87844b9bdc5361cd/filelock-3.4.1-py3-none-any.whl
Collecting importlib-metadata>=4.8.3; python_version < "3.8" (from virtualenv)
Downloading https://files.pythonhosted.org/packages/a0/a1/b153a0a4caf7a7e3f15c2cd56c7702e2cf3d89b1b359d1f1c5e59d68f4ce/importlib_metadata-4.8.3-py3-none-any.whl
Collecting importlib-resources>=5.4; python_version < "3.7" (from virtualenv)
Downloading https://files.pythonhosted.org/packages/24/1b/33e489669a94da3ef4562938cd306e8fa915e13939d7b8277cb5569cb405/importlib_resources-5.4.0-py3-none-any.whl
Collecting zipp>=0.5 (from importlib-metadata>=4.8.3; python_version < "3.8"->virtualenv)
Downloading https://files.pythonhosted.org/packages/bd/df/d4a4974a3e3957fd1c1fa3082366d7fff6e428ddb55f074bf64876f8e8ad/zipp-3.6.0-py3-none-any.whl
Collecting typing-extensions>=3.6.4; python_version < "3.8" (from importlib-metadata>=4.8.3; python_version < "3.8"->virtualenv)
Downloading https://files.pythonhosted.org/packages/45/6b/44f7f8f1e110027cf88956b59f2fad776cca7e1704396d043f89effd3a0e/typing_extensions-4.1.1-py3-none-any.whl
Installing collected packages: platformdirs, distlib, filelock, zipp, typing-extensions, importlib-metadata, importlib-resources, virtualenv
Successfully installed distlib-0.3.6 filelock-3.4.1 importlib-metadata-4.8.3 importlib-resources-5.4.0 platformdirs-2.4.0 typing-extensions-4.1.1 virtualenv-20.16.7 zipp-3.6.0
[alice@c4-dev1 ~]$ which virtualenv
~/.local/bin/virtualenv
[alice@c4-dev1 ~]$ virtualenv --version
virtualenv 20.16.7 from ~/.local/lib/python3.6/site-packages/virtualenv/__init__.py
Start by creating a folder specific to the project you are currently working on. Each project folder will have its own unique set of installed packages. For a project that requires Python 3, do the following (once):
[alice@c4-dev1 ~]$ virtualenv -p python3 my_project
created virtual environment CPython3.6.8.final.0-64 in 3162ms
creator CPython3Posix(dest=~/my_project, clear=False, no_vcs_ignore=False, global=False)
seeder FromAppData(download=False, pip=bundle, setuptools=bundle, wheel=bundle, via=copy, app_data_dir=~/.local/share/virtualenv)
added seed packages: certifi==2022.9.24, charset_normalizer==2.0.12, idna==3.4, panda==0.3.1, pip==21.3.1, python_dateutil==2.8.2, pytz==2022.6, requests==2.27.1, setuptools==59.6.0, six==1.16.0, urllib3==1.26.12, wheel==0.37.1
activators BashActivator,CShellActivator,FishActivator,NushellActivator,PowerShellActivator,PythonActivator
Always remember to specify option -p python3 or -p python2 when you call virtualenv. It makes sure that the intended version of Python is used in the virtual environment when python is called.
Now, each time you want to work on your project, go to its folder and active the virtual environment:
[alice@c4-dev1 ~]$ cd my_project
[alice@c4-dev1 my_project]$ . bin/activate ## IMPORTANT! Note period in front
(my_project) [alice@c4-dev1 my_project]$
Note how (my_project) is prepended to the shell prompt when the virtual environment my_project is activate. This tells you that you run in a customized Python environment. Specifically, python3 now points to a local, frozen version:
(my_project) [alice@c4-dev1 my_project]$ which python3
~/my_project/bin/python3
Similarly, python points to:
(my_project) [alice@c4-dev1 my_project]$ which python
~/my_project/bin/python
(my_project) [alice@c4-dev1 my_project]$ python --version
Python 3.6.8
Note how this local python command points to the local python3 command. What is interesting, and important to notice, is that if we set up a Python 3 virtual environment, then the local python command will point to the local python3 command. In other words, when we use virtual environments, the python command will be using either Python 2 or Python3 at our choice.
To see what Python packages are installed in the virtual environment, use:
(my_project) [alice@c4-dev1 my_project]$ python3 -m pip list
Package Version
------------------ ---------
certifi 2022.9.24
charset-normalizer 2.0.12
idna 3.4
numpy 1.19.5
panda 0.3.1
pandas 1.1.5
pip 21.3.1
python-dateutil 2.8.2
pytz 2022.6
requests 2.27.1
setuptools 59.6.0
six 1.16.0
urllib3 1.26.12
wheel 0.37.1
(my_project) [alice@c4-dev1 my_project]$
With a virtual environment enabled, you can install Python packages to the project folder using python3 -m pip install ... without specifying --user. For instance,
(my_project) [alice@c4-dev1 ~]$ python3 -m pip install pandas
Requirement already satisfied: pandas in ./my_project/lib64/python3.6/site-packages (1.1.5)
Requirement already satisfied: pytz>=2017.2 in ./my_project/lib/python3.6/site-packages (from pandas) (2022.6)
Requirement already satisfied: python-dateutil>=2.7.3 in ./my_project/lib/python3.6/site-packages (from pandas) (2.8.2)
Requirement already satisfied: numpy>=1.15.4 in ./my_project/lib64/python3.6/site-packages (from pandas) (1.19.5)
Requirement already satisfied: six>=1.5 in ./my_project/lib/python3.6/site-packages (from python-dateutil>=2.7.3->pandas) (1.16.0)
To see which packages are now installed in the virtual environment (the “project folder”) and what their versions are, do:
(my_project) [alice@c4-dev1 my_project]$ python3 -m pip list
Package Version
------------------ ---------
certifi 2022.9.24
charset-normalizer 2.0.12
idna 3.4
numpy 1.19.5
panda 0.3.1
pandas 1.1.5
pip 21.3.1
python-dateutil 2.8.2
pytz 2022.6
requests 2.27.1
setuptools 59.6.0
six 1.16.0
urllib3 1.26.12
wheel 0.37.1
(my_project) [alice@c4-dev1 my_project]$
Whenever you open a new terminal, make sure to activate the virtual environment (“project folder”), otherwise it will not find the packages you’ve installed. Pay attention to the shell prompt:
[alice@c4-dev1 ~]$ cd my_project
[alice@c4-dev1 my_project]$ . bin/activate ## ACTIVATE
(my_project) [alice@c4-dev1 my_project]$ pip3 show pandas
Name: pandas
Version: 1.1.5
Summary: Powerful data structures for data analysis, time series, and statistics
Home-page: https://pandas.pydata.org
Author:
Author-email:
License: BSD
Location: ~/my_project/lib64/python3.6/site-packages
Requires: numpy, python-dateutil, pytz
Required-by:
(my_project) [alice@c4-dev1 my_project]$
When submitting a job to the scheduler, make sure the job script loads all required modules and activates the virtual environment.
To deactivate a Python virtual environment, either open a fresh terminal (e.g. log out and back in), or use:
(my_project) [alice@c4-dev1 ~]$ deactivate
[alice@c4-dev1 ~]$
Note how prefix (my_project) was dropped from the shell prompt and python3 now points to the system-wide installation;
[alice@c4-dev1 ~]$ which python3
/usr/bin/python3
The version of Python 3 that comes built-in with CentOS is a bit dated. Specifically, it provides:
[alice@c4-dev1 ~]$ python --version
Python 3.6.8
Newer version of Python 3 are available via SCL:s. The easiest way to use them is to load the SCL as a module, e.g.
[alice@c4-dev1 ~]$ python --version
Python 3.6.8
[alice@c4-dev1 ~]$ module load CBI scl-rh-python
[alice@c4-dev1 ~]$ python --version
Python 3.8.13
Note: You need to load this module before you create a virtual environment.
You will at times get warnings that you are running an old version of ‘pip’:
You are using pip version 8.1.2, however version 20.3.3 is available.
You should consider upgrading via the 'pip install --upgrade pip' command.
Don’t use the suggested command call in that message. Instead, use:
[alice@c4-dev1 ~]$ python3 -m pip install --user --upgrade pip
Cache entry deserialization failed, entry ignored
Collecting pip
Downloading https://files.pythonhosted.org/packages/a4/6d/6463d49a933f547439d6b5b98b46af8742cc03ae83543e4d7688c2420f8b/pip-21.3.1-py3-none-any.whl (1.7MB)
Installing collected packages: pip
Successfully installed pip-21.3.1
You are using pip version 21.3.1, however version 22.3.1 is available.
You should consider upgrading via the 'pip install --upgrade pip' command.
Note, if you’re using a virtual environment, drop --user, i.e.
[alice@c4-dev1 ~]$ python3 -m pip install --upgrade pip
To check the installed version of the ‘pip’ module, use:
[alice@c4-dev1 ~]$ python3 -m pip --version
pip 21.3.1 from ~/.local/lib/python3.6/site-packages/pip (python 3.6)
Python 2 reached the end of its life on 2020-01-01 in favor of Python 3. At this point, the pip maintainers decided to stop making newer versions backward compatible with Python 2. Because of this, Python 2 only supports pip (< 21). To upgrade to the latest supported pip version for Python 2, we can use:
[alice@c4-dev1 ~]$ python2 -m pip install --user --upgrade "pip<21"
DEPRECATION: Python 2.7 reached the end of its life on January 1st, 2020. Please upgrade your Python as Python 2.7 is no longer maintained. pip 21.0 will drop support for Python 2.7 in January 2021. More details about Python 2 support in pip can be found at https://pip.pypa.io/en/latest/development/release-process/#python-2-support pip 21.0 will remove support for this functionality.
Collecting pip<21
Downloading pip-20.3.4-py2.py3-none-any.whl (1.5 MB)
Installing collected packages: pip
Attempting uninstall: pip
Found existing installation: pip 20.3
Uninstalling pip-20.3:
Successfully uninstalled pip-20.3
Successfully installed pip-20.3.4