Virtualenv is a tool used to create an isolated Python environment. This environment has its own installation directories that doesn't share libraries with other virtualenv environments (and optionally doesn't access the globally installed libraries either).
Virtualenv is the easiest and recommended way to configure a custom Python environment.
Difference between virtualenv and venv
venv is a package that comes with Python 3. Python 2 does not contain venv.
virtualenv is a library that offers more functionality than venv. View the following link for a list of features venv does not offer compared to virtualenv.
Although you can create a virtual environment using venv with Python3, it's recommended that you install and use virtualenv instead.
Installing Virtualenv using pip3
Virtualenv is only installed on DreamHost servers for Python 2. If you're working with Python 3, you must install virtualenv using pip3.
pip3 is not installed on the server by default. You must first install a custom version of Python 3. This custom installation of Python 3 will also contain pip3. After it's installed and activated, run the which python3 command as shown in the article. This should return the location of your custom version of Python 3. You should then upgrade pip3.
[server]$ python3 -m pip install --upgrade pip
Once upgraded, install virtualenv using pip3:
[server]$ pip3 install virtualenv
Collecting virtualenv Downloading virtualenv-15.1.0-py2.py3-none-any.whl (1.8MB) 100% |████████████████████████████████| 1.8MB 367kB/s Installing collected packages: virtualenv Successfully installed virtualenv-15.1.0
You'll need the full path to the Python 3 version of virtualenv, so run the following to view it:
[server]$ which virtualenv /home/username/opt/python-3.10.1/bin/virtualenv
Creating a virtual environment using a custom Python version
Be aware that you may need to reinstall Python following a server operating system upgrade.
When working with virtual environments in Python, it's common to use a custom version of Python rather than the server's version. To create a new virtual environment using your custom installed version of Python, follow these steps:
The following steps use Python version 3.10.1. Make sure to use the version you installed.
- Make a note of the full file path to the custom version of Python you just installed. If you've followed the instructions in the installation article, the full path is:
[server]$ which python3 /home/username/opt/python-3.10.1/bin/python
- Navigate to your site's directory, where you'll create the new virtual environment:
[server]$ cd ~/example.com
- Update your .bash_profile
[server]$ . ~/.bash_profile
- Create the virtual environment while you specify the version of Python you wish to use. The following command creates a virtualenv named 'venv' and uses the -p flag to specify the full path to the Python3 version you just installed:
You can name the virtualenv anything you like.
[server]$ virtualenv -p /home/username/opt/python-3.10.1/bin/python3 venv Running virtualenv with interpreter /home/username/opt/python-3.10.1/bin/python3
Using base prefix '/home/username/opt/python-3.10.1'
New python executable in /home/username/example.com/env/bin/python3
Also creating executable in /home/username/example.com/env/bin/python
Installing setuptools, pip, wheel...done.
- This command creates a local copy of your environment specific to this website. While working on this website, you should activate the local environment in order to make sure you're working with the right versions of your tools and packages.
You may see the following error when installing.
setuptools pip failed with error code 1` error
If so, run the following:
[user@localhost]$ pip3 install --upgrade setuptools
Try again and you should be able to install without an error.
- To activate the new virtual environment, run the following:
[server]$ source venv/bin/activate
- To verify the correct Python version, run the following:
[server]$ python -V Python 3.10.1
Any package that you install using pip is now placed in the virtual environments project folder, isolated from the global Python installation.
Deactivating your virtualenv
When finished working in the virtual environment, you can deactivate it by running the following:
- This puts you back into your Shell user's default settings.
Deleting your virtual environment
To delete a virtual environment, simply delete the project folder. Using the previous example, run the following command:
[server]$ rm -rf venv
Installing custom modules
View the following article for information on how to use pip to install Python modules.
Errors creating a virtualenv
You may see the following errors when attempting to create a virtualenv using Python 3.7.
AttributeError: module 'importlib._bootstrap' has no attribute 'SourceFileLoader'
OSError: Command /home/username/venv/bin/python3 -c "import sys, pip; sys...d\"] + sys.argv[1:]))" setuptools pip failed with error code 1
Adding the following line when installing a custom version of OpenSSL to your .bash_profile resolves this.
Use the full path to your custom virtualenv
It's also possible that when you run the virtualenv command, it's using a version outside of your custom installation. Try running this instead to confirm the full path to your Python3 virtualenv.
[server]$ which virtualenv /home/username/opt/python-3.8.0/bin/virtualenv
This should respond with the version in your custom Python3 directory. You can then create a virtualenv using the full path like this:
[server]$ /home/username/opt/python-3.8.0/bin/virtualenv -p /home/username/opt/python-3.8.0/bin/python3 venv