The µ-tutorial on Numpy is released!

Hello!

I am happy to announce that my first µ-tutorial is just being released!

As mentioned two weeks ago, the first topic I chose was Numpy - one of the most powerful Python’s libraries. I know what you might think, it’s plenty of stuff on Numpy out there, so why making the video? Well. Numpy is the kind of library that is always used around whenever you need to process your data. Throughout the years, I have found that despite new libraries and frameworks appear for models, statistics, pipelines, and so on, Numpy is a “Swiss knife” for data. The more natural it feels in your hand, the more efficient you just get with everything…

Now, I have decided to make this tutorial go a bit beyond “micro”, just for the fact that for some of you it may be the first contact with Python (it was for me some years ago!). It’s 2h 20min of a video material divided into six lessons of growing difficulty. If you are already familiar with the basics, you may go to the two last lessons and treat them as exercises. Although, throughout the video I strongly emphasize the need for the so-called vectorization of the code, which is absolutely necessary if you want it to run fast…

In the tutorial, I show how to write understandable and efficient Numpy code. As an example, I show a “toy case” of building a predictive model that can tell the next element of the following sequence: 1, 2, 4, 8, 16, 31, ?, (so expect a little bit of math). After all, you feel you are getting good when you know you can basically turn any equations into code, right?

Here is the link here hope you will enjoy!

Now, for the next steps, I guess it will be a time for writing some articles while I think about the next topic. Ultimately, I want the µ-tutorials to be ~30 minute long videos, but this time I cannot stress that enough how important it is to be fluent with Numpy in pretty much any data project.

As usual, please give me feedback at oleg@zerowithdot, and see you / hear you / read you soon!

Oleg