

Virtual Environments

Using virtual environments is standard practice in Python. A virtual environment is an isolated installation of Python with associated packages. When you use virtual environments, one project can have a separate version of Python and packages. Another project can use a different virtual environment and therefore have a different version of Python and a different set of packages.

Two projects on the same computer can use different versions of Python and different versions of packages if virtual environments are used.

Create a virtual environment with the Anaconda Prompt

To create the new virtual environment, open the Anaconda Prompt and issue the command:

> conda create --name env_name python=3.7

The conda create command builds the new virtual environment. The --name env_name flag gives the new virtual environment the name env_name. Including python=3.7 ensures the virtual environment has a current version of Python.

The following output or something similar results:

The following NEW packages will be INSTALLED:

  ca-certificates    ca-certificates-2019.5.15-0
  certifi            certifi-2019.6.16-py37_0
  openssl            openssl-1.1.1c-he774522_1
  pip                pip-19.1.1-py37_0
  python             python-3.7.3-h8c8aaf0_1
  setuptools         setuptools-41.0.1-py37_0
  sqlite             sqlite-3.28.0-he774522_0
  vc                 vc-14.1-h0510ff6_4
  vs2015_runtime     vs2015_runtime-14.15.26706-h3a45250_4
  wheel              wheel-0.33.4-py37_0
  wincertstore       wincertstore-0.2-py37_0

Type y to confirm and create the new virtual environment.

Activate a virtual environment

To use the new virtual environment env_name, the virtual environment needs to be activated. To activate the environment env_name, issue the command:

> conda activate env_name

The virtual environment is active when you see (env_name) > in parenthesis at the start of the prompt.

(env_name) >

Install packages in a virtual environment

When a new virtual environment is created, no packages are installed by default. If you use the Anaconda distribution of Python, the base environment contains about 600 packages that come with Anaconda. But a fresh new virtual environment will just have Python installed, no other packages.

To install a package into the virtual environment env_name, first make sure the environment is active ((env_name) before the prompt). Then package installation is accomplished with the conda install command followed by the package name. To install Matplotlib into a virtual environment type:

(env_name) > conda install matplotlib

Multiple packages can be installed with the same command. To install both NumPy and Jupyter at the same time use:

(env_name) > conda install numpy jupyter

Deactivate a virtual environment

To deactivate an active environment, run the command:

(env_name) > conda deactivate
>

When the virtual environment is deactivated, the prompt looks normal (just >), with no environment name in parenthesis before it.

List your virtual environments

View a list of your virtual environments using the command conda info --envs or conda env list.

> conda activate env_name
(env_name) > conda info --envs

# conda environments:
#
base                     /home/tribilium/anaconda3
env_name               * /home/tribilium/anaconda3/envs/env_name
matplotlib               /home/tribilium/anaconda3/envs/matplotlib

Notice the * asterisk on the line with env_name. The virtual environment with the * is currently active.

To exit the virtual environment, use the command conda deactivate.

(env_name) > conda deactive

If you run conda info --envs again, there is no * in front of env_name. That's because the env_name virtual environment is no longer active.

> conda info --envs

# conda environments:
#
base                  *  /home/tribilium/anaconda3
env_name                 /home/tribilium/anaconda3/envs/env_name
matplotlib               /home/tribilium/anaconda3/envs/matplotlib

Remove a virtual environment

Remove a virtual environment with the command conda remove --name env_name --all. You need to exit out of the environment before you remove it.

Check to see if the virtual environment you want to remove is present:

> conda info --envs

# conda environments:
#
base                  *  /home/tribilium/anaconda3
env_name                 /home/tribilium/anaconda3/envs/env_name
matplotlib               /home/tribilium/anaconda3/envs/matplotlib

Remove the virtual environment called env_name using the command.

> conda remove --name env_name --all

Proceed ([y]/n)?

Type [y] to remove the environment. Now view a list of your virtual environments again. Note that env_name is no longer in the list.

> conda info --envs

# conda environments:
#
base                  *  /home/tribilium/anaconda3
matplotlib               /home/tribilium/anaconda3/envs/matplotlib