Top all-time · #5853
Cameron Davidson-Pilon
Personal blog · dataorigami.net/blogs/napkin-foldingdataprogramming
Cameron Davidson-Pilon writes Data Origami, a personal blog at dataorigami.net about data and programming. Since 2014, 14 different Hacker News readers have linked it 20 times, most often “Data Science Screencasts”. The last post was in 2020, but the archive is still widely read.
Too few sampled comments (8) to say how often HN recommends it.?
Why HN links it
HN readers link Cameron Davidson-Pilon's blog for Bayesian statistics and probabilistic programming, including the napkin-folding posts on bandits and Optimizely's testing, and a Bayesian book recommendation.
Summarised by Claude from a sample of 8 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Data science, statistics, Python and machine learning blog by Cameron Davidson-Pilon
Most-linked posts?
- /Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/3 people · 3 mentions · first linked 2023-06-21
- Data Science Screencasts2 people · 2 mentions · first linked 2014-06-25
- Bayesian Bandits: Beat A/B testing everytime2 people · 2 mentions · first linked 2017-05-28
What Hacker News says?
Point taken, but there is a much more fundamental issue with it - and precisely why I wrote "very conservative". It is a different problem if we pick two sets from the same data distribution, A and B, and first we have a score on A, then on B. Here we re-run on precisely the same set of Terminal Bench 2.1 problems. It may be that results are so random between runs that each single task has the same probability in a…
Comment linking /Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/
on “Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses”Submitted MapReduce Joins, implemented in PySpark as a story · 4 points
HN discussionFitting a binary classifier on failure doesn't seem quite right to me. There are some quite natural questions you can't answer unless you can predict some stats on when failure will occur, and if you have any drift in the dataset, you might just find features correlated with age (and hence having more time to fail) that don't inherently increase risk. I would recommend: (a) using survival analysis https://lifelines.…
Comment linking /blogs/napkin-folding/18308867-video-presentation-on-lifelines-survival-analysis-in-python-sept-23-2014
on “Predicting Hard Drive Failure with Machine Learning”Submitted No, your A/B test did not increase conversions by 200% as a story · 4 points
HN discussionIt really is not; Bayesian statistics is way more intuitive. For a practical guide see https://dataorigami.net/Probabilistic-Programming-and-Bayesi... For Bayesian ML see https://probml.github.io/pml-book/book1.html and https://www.bishopbook.com/ For Bayesian statistics see http://www.stat.columbia.edu/~gelman/book/
Comment linking /Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/
on “Bayesians moving from defense to offense”Submitted Bayesian Bandits: Beat A/B testing everytime as a story · 3 points
HN discussionThere's a video (now pay-walled) from Cameron Davidson Pilon of Bayesian Methods for Hackers fame, where he takes this approach to build a basic Markov Chain model to predict patterns in his browsing. https://dataorigami.net/collections/all/products/create-mark...
Comment linking /collections/all/products/create-markov-chains-using-your-chrome-browsing-history
on “Analyzing Your Browser History Using Python and Pandas”Submitted 8 great data blogs to follow as a story · 3 points
HN discussionHere's a great article about how Optimizely gets it wrong: http://dataorigami.net/blogs/napkin-folding/17543303-the-bin... There are _many_ offenders. I've yet to see a commercial tool that gets it right. Tragically, the revamp by Optimizely neglects the straightforward Bayesian solution and uses a more fragile and complex sequential technique.
Comment linking /blogs/napkin-folding/17543303-the-binary-problem-and-the-continuous-problem-in-a-b-testing
on “How Optimizely Almost Got Me Fired (2014)”Submitted Data Science Screencasts as a story · 3 points
HN discussionI would strongly recommend the post he cited. It is the same style but features interactive visualizations: https://dataorigami.net/blogs/napkin-folding/79031811-multi-... I implemented something like this for my company and found the latter article quite helpful in explaining the concept to people who understood the basics of probability but not programming.
Comment linking /blogs/napkin-folding/79031811-multi-armed-bandits
on “Multi-Armed Bandits, Conjugate Models and Bayesian Reinforcement Learning”Submitted The Class Imbalance Problem in A/B Testing as a story · 2 points
HN discussionIf you're looking for some Bayesian Stats books that require very little math/stats, I'd recommend starting with these two: Think Bayes by Allen Downey: https://allendowney.github.io/ThinkBayes2/ Bayesian Methods for Hackers by Cam Davidson-Pilon: https://dataorigami.net/Probabilistic-Programming-and-Bayesi... They're both an excellent warm up for Statistical Rethinking.
Comment linking /Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/
on “A Bayesian View of Amazon Resellers”Submitted FiveThirtyEight plot styles in Python as a story · 2 points
HN discussionTwo thoughts. 1. There is a vast literature in "Bayesian Change Point Detection." So you might find a point where something changed and a string of negative reviews began. https://dataorigami.net/Probabilistic-Programming-and-Bayesi... 2. Similarly, there a bajillion ways to weight recent data. One way is to increase the gain on a kalman filter. That will make recent observations more important. There are bayesian i…
Comment linking /Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/
on “A Bayesian View of Amazon Resellers”Submitted Data Origami – Data Science Screencasts as a story · 2 points
HN discussionSubmitted V0.25.0 lifelines released – Python hazard modeling library as a story · 1 points
HN discussionSubmitted Percentile and Quantile Estimation of Big Data: The T-Digest as a story · 1 points
HN discussionSubmitted The Future of Data as a story · 1 points
HN discussionSubmitted ShowHN: Data Origami Screencasts – level up your data science skills as a story · 1 points
HN discussion
