Quickest way to try out Jupyter Notebook: zero install, 3 CLI commands and 5 minutes to action

Lucas Jellema 2

imageThis brief article shows you the quickest way to trying out Jupyter Notebooks. It will not try to persuade you that you should try them out. You probably know that already. It will not show a complex Notebook in detail – plenty of those are available. It simply tells you:

    • open your browser at the Katacoda Ubuntu 20.4 Playground:

https://www.katacoda.com/courses/ubuntu/playground2004

  • create directory
    mkdir work
  • change into directory
    cd work
  • clone a GitHub Repository:
    git clone https://github.com/lucasjellema/DOAG2020-DataWranglingWithJupyterNotebooks
  • run a Docker container with JupyterLab and Python 3 inside
    docker run --rm -p 10000:8888 -v "$PWD":/home/jovyan/work -e JUPYTER_ENABLE_LAB=yes jupyter/minimal-notebook


    image

  • wait for several minutes while the fairly large container image is pulled, then started
  • when the container is running, the terminal window shows a token. Copy the token value to the clipboard
  • open a terminal for Host 1 at port 10000
    a new browser tab is started and the JupyterLab environment is available
  • open the HelloWorldNotebook and read and execute it. It will introduce you to the concept of the Notebook, to the Pandas en Matplotlib libraries and to some plain old Python.

In less than 10 minutes, you will have had you first hands on experience with JupyterLab, notebooks and Pandas.  Note: the Katacoda environment is available for 60 minutes, then it will die on you, so do not get attached.

And now the same steps again, with screenshots:

    • open your browser at the Katacoda Ubuntu 20.4 Playground:

https://www.katacoda.com/courses/ubuntu/playground2004


image

  • create directory
    mkdir work
  • change into directory
    cd work
  • clone a GitHub Repository:
    git clone https://github.com/lucasjellema/DOAG2020-DataWranglingWithJupyterNotebooks
  • run a Docker container with JupyterLab and Python 3 inside
    docker run --rm -p 10000:8888 -v "$PWD":/home/jovyan/work  -e JUPYTER_ENABLE_LAB=yes jupyter/minimal-notebook


    image

  • wait for several minutes while the fairly large container image is pulled, then started
    image
  • when the container is running, the terminal window shows a token. Copy the token value to the clipboard
  • open a terminal for Host 1 at port 10000

    image


    image

    a new browser tab is started; paste the token

    image

    and the JupyterLab environment is available
    image
  • open the HelloWorldNotebook

    image

    and read and execute it. It will introduce you to the concept of the Notebook, to the Pandas en Matplotlib libraries and to some plain old Python.
    image

Resources

My GitHub Repo with sources and an introductory presentation on Jupyter Notebooks: https://github.com/lucasjellema/DOAG2020-DataWranglingWithJupyterNotebooks

A YouTube recording of my recent session for DOAG2020 on Jupyter Notebooks: https://youtu.be/pT62F542uhM

2 thoughts on “Quickest way to try out Jupyter Notebook: zero install, 3 CLI commands and 5 minutes to action

    1. Hi Grzegorz,
      Thanks the link to Colab – I was not aware of it. It certainly seems a nice environment for experiements – and even easier to start using.

      Anyone can start working on my example notebook using this URL:
      https://colab.research.google.com/github/lucasjellema/DOAG2020-DataWranglingWithJupyterNotebooks/blob/main/HelloWorldNotebook/HelloWorldNotebook.ipynb – opening my notebook no GitHub in the colab environment. Note: the emp.csv file needs to be loaded from the raw github url now: https://raw.githubusercontent.com/lucasjellema/DOAG2020-DataWranglingWithJupyterNotebooks/main/HelloWorldNotebook/emp.csv instead of the local file.

      kind regards,
      Lucas

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.