virtualenvs.md

Pipenv & Virtual Environments

The following steps come from The Hitchhiker's Guide.

Assume all instructions are for python 3.x version.

NOTE pipenv and virtualenv are indepenent of each other. I personally prefer virtualenv because of how smooth it can be with virtualenvwrapper.

Pipenv

1. Ensure python is installed by running this at command prompt or Anaconda: python --version

2. Ensure pip is available: pip --version

3. Install dependency manager for your projects: pip install --user pipenv

4. When ready to install packages for your project:

   • Change into your project directory: cd DIRECTORY
   • Run: pipenv install MODULENAME (where MODULENAME is something like requests)

   This installs the library & creates a Pipfile (used to track dependencies in case you need to reinstall them).

5. When looking to run your script: pipenv run python FILENAME.py

   Using pipenv run ensures your installed packages are available to your script.

virtualenv

More on this was covered in Virtual Environments.

Basically virtualenv is a tool to create isolated Python environments. It creates a folder that has all the necessary executables to use the packages your project would need.

It can also be used standalone - in place of Pipenv!

1. Install via pip: pip install virtualenv

2. Test your installation:virtualenv --version

3. Change into the directory of your project: cd DIRECTORY

4. Create a virtual environment: virtualenv ENVNAME

   If you omit the name of the environment, this will not create a folder for all the executable files & copy of your pip library. It will just put them in the current folder.

   NOTE: conventional name is to use venv as it is readily available in ignore files.

For fancier commands, check out the original MD file.

5. Activate your virtual environment
   • Linux: source VENVNAME/bin/activate
   • PC: \venv\Scripts\activate

The name of the current virtual environment will now appear on the left of the prompt to let you know that it’s active. From now on, any package that you install using pip will be placed in the venv folder, isolated from the global Python installation.

6. Install packages as normal

7. When you have your packages, it is best practice to "freeze" the current state of environment packages to a file: pip freeze > requirements.txt

   You can see the list by running: pip list

   You are able to install the requirements later by running: pip install -r requirements.txt

8. Be sure to add your venv to the gitignore list.

9. When done with venv: deactivate

10. To delete you just need to delete the folder of the venv: rm -rf NAME

virtualenvwrapper

virtualenvwrapper provides tab-completion on environment names. It really helps when you have a lot of environments and have trouble remembering their names.

workon also deactivates whatever environment you are currently in, so you can quickly switch between environments.

Setup

1. Ensure that python, pip, and virtualenv are installed.

2. Install the wrapper (full instructions here)

   • linux:

     $ pip install virtualenvwrapper
     $ export WORKON_HOME=~/Envs
     $ source /usr/local/bin/virtualenvwrapper.sh
     

   • PC (default PATH for WORKON_HOME is %USERPROFILE%\Envs): pip install virtualenvwrapper-win

Usage - Create & Work On

I will provide two options - and highly suggest the 2nd.

Standard - Base Python Install

1. Create a venv: mkvirtualenv VENVNAME

   This creates the VENVNAME folder inside ~/Envs.

2. Work on the virtual environment: workon VENVNAME

Standard - Specific Python Install (Not Base)

Maybe you have multiple python versions on your machine - what then?

It's very simple. You will use this command instead: mkvirtualenv -p "\Folder Of\PythonVER\python.exe" VENVNAME

Unless you changed the name of the python executable, that's basically the gist. And you can still create a project!

Alternative

You can make a project, which creates the virtual environment, and also a project directory inside $WORKON_HOME, which is cd-ed into when you workon PROJECTNAME.

1. Create project: mkproject PROJECTNAME

   The first time you run this, you will see a similar message:

   (base) C:\Users\Kat\Documents\SOME_DIRECTORY>mkvirtualenv xlsxwriter-autofit
   C:\Users\Kat\Envs is not a directory, creating
   Using base prefix 'c:\\programdata\\anaconda3'
   New python executable in C:\Users\Kat\Envs\xlsxwriter-autofit\Scripts\python.exe
   Installing setuptools, pip, wheel...
   done.
   
   (xlsxwriter-autofit0) (base) C:\Users\Kat\Documents\Programming\repo\Personal-Projects\xlsxwriter-autofit>
   

   This is because prior to this, you did not yet have the Envs folder that is used to store these virtual environments.

   Since you do not have to be in the root of your project folder to do this, you can run while in any directory. This will (by default) also activate the environment.

2. Set your project directory by creating or navigating to the root folder of your project and then run: setprojectdir

   When you set your project directory, the next time you activate your virtual environment it will:

   • activate the virtual environment
   • automatically move into this directory

3. To deactivate: deactivate

You can activate your virtual project to work on things again by running: workon PROJECTNAME

Usage - Delete

To delete a virtual environment, run: rmvirtualenv venv

Not sure how to delete a virtual project, since the following doesn't work: rmproject venv

Additional items can be found here.

NOTE: If you delete a venv, you can recreate it and then tie it & it's project together in another folder.

Usage - Duplicate A Virtual Environment

There may come a time where you wish to duplicate a virtual environment.

Thanks to this medium post here's how.

cpvirtualenv ENVNAME TARGETENVNAME

Where Else Can I Learn About virtualenvwrapper

By reading the docs.

I Got An Error About PROJECT_HOME Not Being Setup!

Likely you need to set up your environment variables. Checkout the SetEnvVars doc to see the steps I took to resolve on my Windows machine.

Basically this StackOverflow response was probably the only thing that needed to be done.