When you work with python for web development, data analysis or other solution , you need to install packages from pip and sometimes we have conflicts: Web framework X needs one package  version 1.2 and Web framework Y needs the same package but version 1.4. 

One Great solution for that problem is virtual environment – a complete environment with everything you need in one folder tree. You can simply copy it from one computer to another or delete the folder and everything is uninstalled.

Creating a Virtual Environment

Make sure you have python installed (and pip)

Installing virtual environment:

# pip install virtualenv

Creating a new environment :

# virtualenv -p [PATH to python executable] [target directory]

So if for example, python3 interpreter is located in /usr/bin and we want to create a directory myenv for our virtual environment :

# virtualenv -p /usr/bin/python3 myenv

cd into the created directory and you will notice that everything is inside : the interpreter, other tools , needed libraries and header files

In python 3 you can create a virtual environment using the command:

# python -m venv myenv2

Activating the environment:

To activate the environment run the activate script in the current shell:

# source bin/activate

On windows you need to run activate.bat script in the Scripts directory

You will see the virtual env name in your prompt:

(myenv) developer@comp1: ~/myenv$

Note that the PATH variable is changed to include the environment bin directory first

Installing packages in the virtual environment:

You can install any package or tool in your virtual environment using pip. 

Installing Jupyter and ipython:

# pip install ipython
# pip install jupyter

You can find the tools in the bin (scripts in windows) directory

Installing some packages

# pip install numpy 
# pip install tensorflow

You can find the packages in lib/python/site-packages directory

Now you can start jupyter notebook and play with numpy and tensorflow

# jupyter notebook

Saving the virtual environment:

If you find create many virtual environments with the same packages (different web applications for example) you can save the requirements and install it again easily :

# pip freeze > requirements.txt

Use the file requirements.txt in another environment using a simple command:

# pip install -r requirements.txt

Leaving the virtual environment

To leave the environment:

# deactivate

The bin directory is removed from the PATH variable

Delete the virtual environment

Simply delete the directory