[python] Use virtualenv with Python with Visual Studio Code in Ubuntu

I have a Python project and navigation/autocomplete work perfectly within files in the project. I want to specify a virtualenv so that navigation/autocomplete work with packages inside that virtualenv.

I tried this in settings.py, but navigation/autocomplete does not work. Also by setting "python.pythonPath":"~/dev/venvs/proj/bin/python killed the original navigation/autocomplete.

{
    "editor.rulers": [80,100],
    "python.autoComplete.extraPaths": [
        "~/dev/venvs/proj",
        "~/dev/venvs/proj/lib"
     ]
}

I was able to use the workspace setting that other people on this page have been asking for.

In Preferences, ?+P, search for python.pythonPath in the search bar.

You should see something like:

// Path to Python, you can use a custom version of Python by modifying this setting to include the full path.
"python.pythonPath": "python"

Then click on the WORKSPACE SETTINGS tab on the right side of the window. This will make it so the setting is only applicable to the workspace you're in.

Afterwards, click on the pencil icon next to "python.pythonPath". This should copy the setting over the workspace settings.

Change the value to something like:

"python.pythonPath": "${workspaceFolder}/venv"

On Mac OS X using Visual Studio Code version 1.34.0 (1.34.0) I had to do the following to get Visual Studio Code to recognise the virtual environments:

Location of my virtual environment (named ml_venv):

/Users/auser/.pyvenv/ml_venv

auser@HOST:~/.pyvenv$ tree -d -L 2
.
+-- ml_venv
    +-- bin
    +-- include
    +-- lib

I added the following entry in Settings.json: "python.venvPath": "/Users/auser/.pyvenv"

I restarted the IDE, and now I could see the interpreter from my virtual environment:

Another way is to open Visual Studio Code from a terminal with the virtualenv set and need to perform F1 Python: Select Interpreter and select the required virtualenv.

Quite simple with the latest version of Visual Studio Code, if you have installed the official Python extension for Visual Studio Code:

Shift + Command + P

Type: Python: Select Interpreter

Choose your virtual environment.

Official site

It seems to be (as of 2018.03) in code-insider. A directive has been introduced called python.venvFolders:

  "python.venvFolders": [
    "envs",
    ".pyenv",
    ".direnv"
  ],

All you need is to add your virtualenv folder name.

I put the absolute path of the virtual environment Python executable as well has the packages. I then restarted Visual Studio Code.

I am trying to get ${workspaceRoot} to avoid hardcoding absolute paths.

{
    "editor.rulers": [80,100],
    "python.pythonPath": "/home/jesvin/dev/ala/venv/bin/python",
    "python.autoComplete.extraPaths": [
        "/home/jesvin/dev/ala/venv/lib/python2.7",
        "/home/jesvin/dev/ala/venv/lib/python2.7/site-packages"
     ]
}

I got this from YouTube Setting up Python Visual Studio Code... Venv

OK, the video really didn't help me all that much, but... the first comment under (by the person who posted the video) makes a lot of sense and is pure gold.

Basically, open up Visual Studio Code' built-in Terminal. Then source <your path>/activate.sh, the usual way you choose a venv from the command line. I have a predefined Bash function to find & launch the right script file and that worked just fine.

Quoting that YouTube comment directly (all credit to aneuris ap):

(you really only need steps 5-7)

1. Open your command line/terminal and type `pip virtualenv`.
2. Create a folder in which the virtualenv will be placed in.
3. 'cd' to the script folder in the virtualenv and run activate.bat (CMD).
4. Deactivate to turn of the virtualenv (CMD).
5. Open the project in Visual Studio Code and use its built-in terminal to 'cd' to the script folder in you virtualenv.
6. Type source activates (in Visual Studio Code I use the Git terminal).
7. Deactivate to turn off the virtualenv.

As you may notice, he's talking about activate.bat. So, if it works for me on a Mac, and it works on Windows too, chances are it's pretty robust and portable.

With the latest Python extension for Visual Studio Code, there is a venvPath Setting:

// Path to folder with a list of Virtual Environments (e.g. ~/.pyenv, ~/Envs, ~/.virtualenvs).
  "python.venvPath": "",

On Mac OS X, go to Code ? Preferences ? Settings and scroll down to Python Configuration.

Look for "python.venvPath: "", and click the pencil on the left-hand side to open up your user settings. Finally, add the path to where you store your virtual environments.

If you are using virtuanenvwrapper, or you have put all your virtual environment setting in one folder, this will be the one for you.

After you have configured "python.venvPath", restart Visual Studio Code. Then open the command palette and look for "Python: Select Interpreter". At this point, you should see the interpreter associated with the virtual environment you just added.

a) Modify Visual Studio Code default virtual env path setting. It's called "python.venvPath". You do this by going into code->settings and scroll down for python settings.

b) Restart VS Code

c) Now if you do Shift + Command + P and type Python: Select Interpreter you should see list of your virtual environments.

As of September 2016 (according to the GitHub repository documentation of the extension) you can just execute a command from within Visual Studio Code that will let you select the interpreter from an automatically generated list of known interpreters (including the one in your project's virtual environment).

How can I use this feature?

• Select the command Python: Select Workspace Interpreter(*) from the command palette (F1).

• Upon selecting the above command a list of discovered interpreters will be displayed in a quick pick list.

• Selecting an interpreter from this list will update the settings.json file automatically.

(*) This command has been updated to Python: Select Interpreter in the latest release of Visual Studio Code (thanks @nngeek).

Also, notice that your selected interpreter will be shown at the left side of the statusbar, e.g., Python 3.6 64-bit. This is a button you can click to trigger the Select Interpreter feature.