Data Science from Scratch: First Principles with Python (2nd Edition) Author: Joel Grus

Data Science from Scratch: First Principles with Python (2nd Edition) by Joel Grus offers a refreshing dive into the essence of data science through a bottom-up approach. Perfect for curious minds and aspiring analysts, this book builds everything from the ground up—focusing not just on using tools, but truly understanding how they work.

Grus begins by unpacking the fundamental building blocks of data science: basic statistics, linear algebra, probability, and Python programming. With clarity and wit, he guides readers through essential tasks like data manipulation, visualization, and machine learning—with no reliance on heavy libraries or black-box methods.

Key highlights include:

Crafting algorithms with pure Python

Digging into statistics and probability for analytics

Demystifying machine learning techniques like decision trees, k-nearest neighbors, and neural networks

Building your own models and tools to understand how data science solutions work beneath the surface

Encouraging critical thinking and experimentation over blindly following packaged solutions

The book isn’t just educational—it’s empowering. Ideal for students, developers pivoting into data science, or self-taught enthusiasts, it gives readers the confidence to build, debug, and improve their own analytical tools. Published by O'Reilly Media, it’s part of a trusted series known for combining practical learning with deep technical insight.