JAKE VANDERPLAS: Hi, and welcome to this video series

on Google Colab.

I'm Jake VanderPlas, and I'll be your guide today

as we look at what Colab is all about.

Google Colab is an executable document

that lets you write, run, and share code within Google Drive.

If you're familiar with the popular Jupyter project,

you can think of Colab as a Jupyter notebook

stored in Google Drive.

A notebook document is composed of cells, each of which

can contain code, text, images, and more.

Colab connects your notebook to a cloud-based runtime,

meaning you can execute Python code without any required

setup on your own machine.

Additional code cells are executed

using that same runtime, resulting

in a rich, interactive coding experience in which you

can use any of the functionality that Python offers.

For example, here we define a variable containing

a range of 10 numbers.

In the next cell, we loop through this range,

printing the square of each number.

For convenience, we use the Shift-Enter shortcut

rather than the Play button to execute the cell.

Cell outputs are not limited to simple text, however.

They can contain any number of dynamic, rich outputs.

For example, we can search Colab's built-in library

of code snippets and insert code to create an interactive data

visualization.

This particular visualization is created

with Altair, one of several third-party visualization

libraries that Colab supports.

Colab notebooks can be shared like a Google Doc,

and for this purpose it's useful to use

text cells to provide a narrative around the code

you've executed.

Text cells are formatted using Markdown, a plain text document

format that's rendered on the page.

Markdown format is simple and powerful,

allowing you to add headings, paragraphs, lists,

and even mathematical formulae.

If you would like to share your notebooks with others,

you can do so via Google Drive sharing

or even by exporting your notebook to GitHub.

The notebook is stored in the standard Jupyter Notebook

format, and so the notebooks you create

can be viewed and executed in Jupyter Notebook, JupyterLab,

and other compatible frameworks.

The convenience of sharing notebooks

means that you can find and explore

many interesting notebooks around the web.

One useful collection is the Seedbank project

at research.google.com/seedbank.

For example, the Neural Style Transfer seed

shows how to use deep learning to transfer styles

between images and includes a link

to a Colab notebook where you can run and modify the code.

To learn more about Colab, visit colab.research.google.com

and find the Welcome notebook, where

you will find links to tutorials and other info

about Jupyter and Colab notebooks.

You can also find the remaining videos

in this series, which will explore Colab in more depth.

In the next video, my colleague Lawrence

will explore how to install TensorFlow using Colab

and how to use different runtimes

to access things like the GPU.

See you there.

Hi, I'm Jake.

I'm a software engineer on the Google Colab project,

and we've got lots of great videos for you about Colab.

So feel free to hit that Subscribe button.