Give Google Colab a shot

This is going to sound like an ad but I promise you it’s not, it’s just a testimony. I’m genuinely surprised to not see any mention of Google Colab on these forums. Colab is a Jupyter notebook environment. Now, the time I discovered Colab was also the first time I heard about Jupyter, so my simplistic explanation is that it’s like a Google doc that lets you run Python (and Swift, but right now you can only do it by making a copy of an existing notebook running the Swift kernel). This has a lot of benefits like automatic saving, revision history (with diffs!) and collaboration with other people. It also appears to run perfectly fine in Safari on iPad and iPhone.

Over the last few months, Colab has become my IDE an overwhelming majority of the time, in turn making Python my primary language. It’s not my favorite IDE—that would be Visual Studio—but it enables incredibly useful things. Its primary benefit to me is in setting up utilities I know will just run anywhere, both for myself and for other people… especially non-technical ones. For the latter group of people, you can choose the option to hide the code in a notebook and just tell them to hit the play button.

It comes installed with a number of common libraries like pandas, but you can always install your own with a simple !pip install --quiet package-name line in the notebook. Make sure you RTFM for a lot of handy tips. Google has been using Colab extensively in its documentation for TensorFlow to demonstrate code if you’d like examples.

Thought I’d share if anyone would benefit from this!

1 Like

Don’t get me wrong, i love the fact you found something that works for you. However Google isn’t primarily known for keeping systems and on going projects.

In the last years they canceled and stopped existing services that’s what’s keeping me from using Google services.

Google’s business is to keep tabs on everything that is happening in the world. This service has been provided in hopes they can tap into new Machine Learning projects as they are developing. As with so many other Google projects, I expect that once their Machine Learning algorithms are properly trained, they will kill it.

That’s fair! Either way, it works now and there are alternatives if they put a stop to it, since it is in principle just Google’s spin on Jupyter.

Edit: Just adding that some current alternatives are listed on the website.