Final Thoughts on Python and R for the Modern Data Scientist
So we made it through the whole book. And honestly? It was worth the ride.
What This Book Got Right
The biggest thing Scavetta and Angelov got right is the framing. They didn’t write a “Python is better” or “R is better” book. They wrote a “both are useful, here’s when to use which” book. And that’s the mature take.
In my 20+ years working with data and code, I’ve seen so many people get stuck in one tool. They learn Python and try to force it into every situation. Or they learn R and refuse to touch anything else. But the real world doesn’t care about your preferences. It cares about results.
The Key Takeaways
If I had to boil the whole book down to a few points, here’s what I’d keep:
1. R wins at statistics and visualization. The tidyverse is genuinely good. ggplot2 makes beautiful charts with less code. R Markdown for reports is still hard to beat. If your work is mostly exploratory data analysis and statistical modeling, R is your friend.
2. Python wins at everything else. Web scraping, deployment, deep learning, text processing, APIs. Python’s general-purpose nature means it plugs into more things. If you’re building production systems or need to go beyond pure analytics, Python is the better choice.
3. You don’t have to choose. Tools like reticulate and rpy2 let you mix both languages in the same project. Apache Arrow and Feather give you shared data formats. The walls between Python and R are thinner than ever.
4. The “language war” was always silly. Both languages grew from different needs. R came from statistics. Python came from general programming. They were never really competing. They were solving different problems.
5. Modern data science is about being flexible. The best data scientists pick the right tool for the job. Sometimes that’s R. Sometimes that’s Python. Sometimes it’s both.
What’s Changed Since 2021
The book came out in 2021, and some things have shifted since then. Python’s ecosystem keeps growing fast, especially around machine learning and AI. R’s community is still strong but more focused on its core strengths in statistics and bioinformatics.
The interoperability story has only gotten better. Quarto replaced R Markdown as the go-to for scientific publishing and it works with both languages. Polars is shaking up the data frame world in Python. And the Posit team (formerly RStudio) keeps building bridges between the two ecosystems.
My Personal Take
I’ve used both languages for years. My typical workflow is: explore data in R because ggplot2 is just nicer for quick plots, then build the production pipeline in Python because it deploys easier.
Is that everyone’s workflow? No. But the point is having options.
If you only know one language, this book is a solid guide for picking up the other. If you know both, it’s still useful for understanding how to combine them effectively. Either way, Scavetta and Angelov made a good case that the future of data science is bilingual.
The Full Series
If you missed any posts in this retelling, here’s the complete list:
- Series Introduction
- The Preface - What Modern Data Science Means
- Chapter 1 - Origins of Python and R
- Chapter 2 Part 1 - R for Pythonistas: Getting Started
- Chapter 2 Part 2 - R for Pythonistas: Data Wrangling
- Chapter 3 - Python for R Users
- Chapter 4 - Data Format Context
- Chapter 5 - Workflow Context
- Chapter 6 - Using Both Languages Together
- Chapter 7 - Bilingual Case Study
- Python and R Cheat Sheet
Thanks for reading along. If you found this useful, grab the full book for all the code examples and details I couldn’t fit into blog posts.
Book: Python and R for the Modern Data Scientist by Rick J. Scavetta and Boyan Angelov (O’Reilly, 2021, ISBN: 978-1-492-09340-4)