Avoid the most dangerous pitfall of histograms — Binning Bias, The Biggest Flaw of Histograms Histograms are probably the first plot you might have used as you embark on your journey as a data scientist. They are intuitive and easy plots to understand the shapes of distributions. However, as you progress through your journey, you will find that histograms are not so peachy. Histograms group…

Use Pandas the way it was intended to — Introduction “Great… another article on how to make Pandas n times faster.” I think I have said that countless times for the past three years I have been using Pandas. The most recent one I saw said, “make Pandas 71,803 times faster”. But I won’t give you that kind of promise…

No error messages — that’s what makes them subtle — Introduction We are all used to the big, fat, red error messages that frequently pop up while we code. Fortunately, people won’t spot it because we always fix those errors. But how about the mistakes that give no errors? These are the trickiest, but the pros could easily call them out. …

Transform and manipulate images at will with scikit-image — Introduction We are here on a sad business. Very sad, indeed. We are here to learn how to take beautiful, breathtaking images and turn them into a bunch of ugly little numbers so that they are more presentable to all those soulless, mindless machines. We will take animals and strip them…

Use Matplotlib like never before — I used to have a rule… When I was a beginner learning data visualization, I used to have a rule — never, ever visit the Matplotlib documentation. Nothing good ever came out of there. When I was on its pages, my eyes would process English words but, somehow, they were interpreted into my brain like I…

Instead of 1000 words, use a single striking sketch. — Introduction Even though you’ve heard it countless times, I repeat it — a picture is worth a thousand words. In data science communication, this quote is literally, figuratively, any-possible-adverb-here true. You don’t want to ramble on to your manager, boss or, colleague reading from a wall of text. Instead, deliver your…

Each one is worth your time — Introduction There are two types of tricks in data science and ML: tricks that are rare and very cool. They are designed to grab your attention but ultimately, you will never use them because their use-cases are too narrow. Think of those Python one-liners that are dreadful in terms of readability. …

This is a complete copy/paste of my LinkedIn post I shared last year. At least, give proper credit.

From semi/anti joins to validating data merges — A nasty shock that comes with real-world data is that it never comes in a single, tame CSV file. Instead, it ıs a bundle of tables that interact with each other in many ways using common columns. It is your unfortunate job to leverage these interactions and find insights from…