I am happy to report that the second edition of Think Bayes went to the printer today! Electronic versions will be available in a week or so, print versions in 2-3 weeks.
I’ve been asked to suggest people O’Reilly should send publicity copies to, so I immediately thought of the good people of PyMC.
I’m not sure how many copies O’Reilly is willing to provide (or where they will mail them) so I might not be able to satisfy all requests. But you can always access the free version of the book at https://thinkbayes.com.
If you get a copy, you are under no obligation. But I welcome your comments and suggestions, and if you would like to review it somewhere, that would be great!
I am currently working through “Think Bayes: Bayesian Statistics in Python”
05-18-2021 edition
In Chapter 2, p.24, there is given data, and then a bulleted list of hypotheses and their probability given the data.
The bulleted list is only correct if the contestant has chosen door 1. This assumption is not stated clearly within this section that lays out hypotheses, data, and making it very confusing for a student, especially one not familiar with the Monty Hall Problem.
The text would be better written to reiterate here that door one has been chosen before Monty opens door 3. The probabilities are dependent on which door the contestant has chosen. Unless explicitly stated, the probabilities are actually different than the bulleted list.
For example:
If the contestant did not choose a door at all, but Monty still opened door 3 to a goat (which while against the rules is also possible - chaotic good here) , the probability that the car would be behind door 2 given the data, would be 50% instead of 100% when the contestant chooses door 1.
Thanks for your time. Writing and editing math textbooks is no easy feat!