Introduction to numpy for data science

As a data science student, my coding style as changed a lot over the years. With this post I want to help you to write faster, cleaner and shorter code with the help of numpy.

Arrays in numpy can be sliced and strided using basic indexing like known from python: data[start:end:step]. If you want to want to index starting from the last entry you can use negative values for this indexing. Let me show you a few examples:

Advanced indexing is where it becomes interesting. It helps you modify elements of an array conditionally and working with index arrays for complex selections. This can be done using boolean masks or using index broadcasting.

We want to implement this formula to approximate the square signal for an array of input values t. As the sum over k ranges to infinity, we set it to 200 for this example. It is tempting to use a for loop over k and summing the partial results. Instead, lets have a look at how this can be done without loops:

It makes your code more readable and avoids common errors with the indexes x and y. It still allows you to use the coordinates if you need to draw values from another array for example.

I hope these examples give you a head start on learning python with numpy. Thank you for reading!

Related posts:

An important part of the self-publishing process is accurately filling out the forms on Kindle Direct Publishing, B&N Press, and any other e-book vendors where you’ll upload your book for…

August means peaches here in my world. Every year I visit one of my local farms to buy boxes and boxes of peaches 🍑 It’s my “Annual Peach Pilgrimage” However, last month I took a class to learn how…

Have you ever had a dirty air duct? If not, then you’re probably living in a vacuum cleaner-free world! Dust, pollen, and other allergens accumulate over time in air ducts, which can lead to problems…