Model Blog: David Robinson of Variance Explained

12 Jan 2018 . category: Blogging . Comments

http://varianceexplained.org/

Now this bit of spotlight is long overdue - Variance Explained was my first “model blog” in spirit, if not in post chronology.

David Robinson is so good at blogging. I was recently reminded of this while reading his latest post, “What’s the difference between data science, machine learning, and artificial intelligence?” 1.

His writing is engaging and plain-speaking, and he has a friendly style.

I greatly admire the look and feel of his blog - so much so that I’ve tried replicating his ‎clean way of referencing, linking, and footnoting within posts. His use of these keeps posts streamlined and highly readable.

I even chose to use Jekyll as the static site generator for this blog after noticing it in his footer.

I analyzed and reproduced (swapping Python for R) one of his posts as one of my very first in-depth undertakings for this blog. It introduced me to the concept of tidy data and much more. Other posts I’d recommend:

  1. “Text analysis of Trump’s tweets confirms he writes only the (angrier) Android half” (and its follow-up a year later)
  2. “Don’t teach built-in plotting to beginners (teach ggplot2)” 2
  3. “Two years as a Data Scientist at Stack Overflow”
  4. “Advice to aspiring data scientists: start a blog” (sound familiar?)
  5. “Teach the tidyverse to beginners”

I only discovered Variance Explained this past May, but the rest of his backlog looks just as impressive3.

I am a huge fan and grateful to have David’s continued example.

Footnotes

  1. True to style, it’s an awesome post. “Data science produces insights. Machine learning produces predictions. Artificial intelligence produces actions.” - oversimplification made useful, then followed by a breakdown of what he means by it. I also enjoyed his use of Twitter snark from the community. And the AI effect has its own Wikipedia page! 

  2. This post was actually brought to my attention during the Johns Hopkins R Programming course that I took back in 2016. 

  3. For example, “Understanding mixture models and expectation-maximization (using baseball statistics)” sounds like it could help reinforce some of the stuff I learned about topic models and Latent Dirichlet Allocation. 


Me

Nadja does not particularly enjoy writing about herself.