We will primarily use Jupyter Notebooks to put into practice what we learn throughout this course. Notebooks allow us to work with and run Python code easily, rather than learn a fully-fledged Integrated Development Environment (IDE). This page describes how to install Anaconda and Jupyter Notebooks.
Before we continue, each section containing code on this website also has a link to working code in a Google CoLab live environment. This is functionally the same as Jupyter Notebooks, just in-browser. You can decide if you want to install Anaconda and Jupyter Notebooks. I would recommend it, Colab is good for convenience, however.
Whichever you choose, Jupyter or Colab, do explore the code, consider it and perhaps copy it out again into a new cell. Actively learn what it is doing, rather than just running the code in the cell.
Anaconda bundles Jupyter Notebooks (formerly iPython Notebooks), the Python language, scikit-learn, TensorFlow, matplotlib (and a number of other scientific programs) together for ease of use. We will use each of these throught this course.
- Download the Anaconda installer.
- Launch Anaconda Navigator.
- Launch Juypter Notebook from inside the Anaconda.
Juypter Notebooks in Brief
Juypter Notebook opens in your web browser, running on a server on your local machine. The page that opens is a file browser of sorts, you can navigate and create new notebooks here. New notebooks are created with the plus (+) icon on the top right.
Notebooks organise code into cells, so you can group related code together or write notes and comments above/below code. You can set the type (code/markdown (text)) of cell in a drop-down menu. Code in a cell is run by pressing the play button and stopped by pressing stop. If the kernel ‘dies’ - ceases to function due to an error - you can simply restart it. The status of the kernel is indicated by the circle below the logout button on the right.
Asynchronous Runtime in Notebooks
Juypter Notebooks runs each cell in a single event loop, this means while one cell is running, no other cell can run. You receive a runtime error; “This event loop is already running”. We can apply a small ‘patch’ into a cell to allow you to run multiple event loops within that notebook. Don’t get carried away though, this is to avoid the mild inconvenience of a runtime error, nothing more.
# Asyncio allows multiple event loops to run in Jupyter Notebooks. # Fixes "RuntimeError: This event loop is already running". import nest_asyncio nest_asyncio.apply()