Ubuntu 18 how to install Jupyter Notebook/IPython

John D K -

This is a quick tutorial how to install Jupyter Notebook on Ubuntu 18 or Linux Mint 18.

In this post:

Prerequisite for Jupyter Notebook/IPython

First of all you need to update your cache:

sudo apt-get update
sudo apt-get upgrade

By default you will have installed python and pip. In order to verify their existence and versions:

python --version
pip --version

in case of missing python and/or pip you can install them by:

sudo apt-get -y install python2.7 python-pip python-dev

or for Python 3

sudo apt-get -y install python3.5 python-pip python-dev

I prefer to install Jupyter in virtual environment so you can create one if you don't have byReferences:

virtualenv --system-site-packages -p python3 jupyenv

then go in the newly created environment and activate it by:

cd jupyenv
source ./bin/activate

you should see terminal change like:

(jupyenv) user@user ~/envs/jupyenv $

in order to stop virtual environment type:

deactivate

returning your terminal to normal.

You can see more about virtual environments here: Install python 3.7 on Linux Mint

Install and run Jupyter Notebook / IPython

After ensuring Python, pip and virtual environment you can install IPython by:

sudo apt-get -y install ipython ipython-notebook

Activate your virtual environment and install Jupyter Notebook in this environment by:

pip install jupyter

if you have errors you can try to upgrade pip by:

pip install --upgrade pip

Now you can run it by:

jupyter notebook

If you like you can create starting script for Jupyter Notebook by:

#!/bin/bash
cd /home/user/envs/jupyenv
source /home/user/envs/jupyenv/bin/activate
jupyter notebook

This will start Jupyter which will be available in the browser by opening address:

http://localhost:8888/tree

Testing and running simple Notebook

Inside the Jupyter you can create new Notebook from button New and selecting Python 2 or Python 3. Then you can enter python code and run the cell:

for i in range(1, 10, 2):
    print(i)

and check the output.

There are many options:

  • You can use different Kernel
    • Python 2
    • Python 3
  • Cell Type
    • Code
    • Markdown

You can create cells with markdown like:

# Notebook Jupyter Notebook

This is a simple example showing for loop and range:
```python
for i in range(1, 10, 2):
    print(i)```

You can create many cells and run them all. In case of problems you can restart the Kernel.

Once you finish with your tests then you can stop the Jupyter by CTRL + C. Your files will be persistent between the restarts.

How to stop or check status of Jupyter

In order to stop Jupyter you can use this command:

jupyter notebook stop 8888

You can check for all running instances by this code:

jupyter notebook list

expected result:

Currently running servers:
http://localhost:8892/?token=b54426bbd5c0a0912137a33aabd775d0ec8e :: /home/user/jupy
http://localhost:8891/?token=baf9fb694459600724dbe1205feda976ebd0 :: /home/user/jupy
http://localhost:8890/?token=1fb8cff5ecc1ea1ac3227bff67242d2331bf :: /home/user/jupy
http://localhost:8889/?token=6e2a519b5488f74ebe82fc6c296d338d2fcc :: /home/user/jupy
http://localhost:8888/?token=866a73699fdf74505f4e5b0bdc0832e5d8a5 :: /home/user/jupy

References

You can do much more with IPython than simply running Python code and discplay Markdown. You can run shell commands, set environmental variables and more. You can check some of the tricks on the links below: