What is IPython Notebook


The first, very succinct answer is: It depends. Jupyter can actually mean several things. On the one hand, there is the Jupyter project, an initiative for open source software or open standards and services that focuses on interactive computing with the support of as many programming languages ​​as possible. Project Jupyter offers different products. At the moment these are: Jupyter Notebook, Jupyter Kernels, Jupyter Hub and Jupyter Lab. Whenever the term “Jupyter” comes up, it is mostly Jupyter Notebook meant - or more precisely: The application Jupyter Notebook, with which you can build a digital document, which is also called Jupyter Notebook. In this post we want to deal primarily with the notebook software and the output it produces.

What is special about a Jupyter notebook? It could be called the ultimate digital notebook. Forget wikis and Google Drive and Evernote - Jupyter allows not only the elegant integration of texts, links, images and videos but also the listing of code and its execution in bites on the spot. In other words: A Jupyter notebook can also contain interactive maps or data visualizations - directly behind the corresponding Python snippets. Incidentally, they partly take care of the name of the product, which otherwise goes back to Julia and R: JuPyt-eR.. The notebooks now support dozens of other languages, including heavyweights such as C #, Java, JavaScript, PHP, Ruby, Scala or Swift (this is where the Juptyer kernels come into play).

Jupyter Notebooks were created in the data science community. Their goal, which has now been achieved twice and three times: better documentation, better workflows and better collaboration in the field of scientific computing.

Keyword collaboration: Notebooks can of course be shared in a variety of ways. So you can not only export and send your notebook in the classic way (various formats), but also pack it in a Github repository, where it can then be collaboratively worked on or expanded. If your own Jupyter resources are part of a more complex project that requires additional components for execution, sharing via Docker containers is a good idea.

There are also many options when running the Notebook application. Desktop application, operation in the large cloud (via binder), Jupyter on a self-configured multi-user server (Jupyter Hub) - all no problem.
The next-generation web interface (JupyterLab), which was released at the beginning of 2018, gave the software a further boost.

So Jupyter Notebook is a fantastic tool. Not for data scientists and developers, but also for project managers, journalists and documentaries.

You can find chic notebooks to rummage here, among other places.
If you want to test Jupyter Notebooks in an uncomplicated way, click here.
And the extensive documentation can be found here.