Statistical Rethinking: A Bayesian Course with Examples in R and STAN (Chapman & Hall/CRC Texts in Statistical Science)
Original price was: $100,00.$19,99Current price is: $19,99.
- 100% Satisfaction Guaranteed!
- Immediate Digital Delivery
- Download Risk-Free
✔️ Digital file type(s): 𝟏𝐏𝐃𝐅
Statistical Rethinking: A Bayesian Course with Examples in R and Stan builds your knowledge of and confidence in making inferences from data. Reflecting the need for scripting in today’s model-based statistics, the book pushes you to perform step-by-step calculations that are usually automated. This unique computational approach ensures that you understand enough of the details to make reasonable choices and interpretations in your own modeling work.
The text presents causal inference and generalized linear multilevel models from a simple Bayesian perspective that builds on information theory and maximum entropy. The core material ranges from the basics of regression to advanced multilevel models. It also presents measurement error, missing data, and Gaussian process models for spatial and phylogenetic confounding.
The second edition emphasizes the directed acyclic graph (DAG) approach to causal inference, integrating DAGs into many examples. The new edition also contains new material on the design of prior distributions, splines, ordered categorical predictors, social relations models, cross-validation, importance sampling, instrumental variables, and Hamiltonian Monte Carlo. It ends with an entirely new chapter that goes beyond generalized linear modeling, showing how domain-specific scientific models can be built into statistical analyses.
Features
- Integrates working code into the main text
- Illustrates concepts through worked data analysis examples
- Emphasizes understanding assumptions and how assumptions are reflected in code
- Offers more detailed explanations of the mathematics in optional sections
- Presents examples of using the dagitty R package to analyze causal graphs
- Provides the rethinking R package on the author’s website and on GitHub.
32 reviews for Statistical Rethinking: A Bayesian Course with Examples in R and STAN (Chapman & Hall/CRC Texts in Statistical Science)
You must be logged in to post a review.
Related products
-
Essential Calculus Skills Practice Workbook with Full
Original price was: $33,00.$16,99Current price is: $16,99. -
Proofs from THE BOOK
Original price was: $59,99.$16,99Current price is: $16,99. -
The Math Book (DK Big Ideas)
Original price was: $35,95.$16,99Current price is: $16,99.
TheCozyReader –
Bayesian Statistics was long thought by many outside of statistics to be a mathematician’s game that is impossible to win without a PhD or a lot of time on your hands. You can’t competently program in Stan if you don’t understand Bayesian Inference and you can’t really understand Bayesian Inference if you don’t practice it, so frustration was always the norm for those who wanted to understand this segment of statistics. Online courses tried their best to fill the void, but pretty much none of the online courses I have tried are really any good. On the other hand, rigorous books like Andrew Gelman’s Bayesian Data Analysis are overly-theoretical slogs that will drain the under-prepared of their will power. Statistical Rethinking is the only resource I have ever read that could successfully bring non-Bayesians of a lower mathematical maturity into the fold.
You will actually get to practice Bayesian statistics while learning about it and the book is incredibly easy to follow. You will not become an expert at Bayesian Statistics from this book, but you will have an actual starting point from which you can approach the more difficult texts. I cannot recommend this book enough for beginners.
John Nedson –
I have read and used BDA3 by Gelman et al. and thought I would not read another Bayesian analysis book. But this book is like a romantic Bayesian novel — reading every page makes me want to read the next… It’s an awesome book and I recommend it to anyone interested in the beautiful Bayes’ world!
gianmarco –
Libro consigliatissimo per ricercatori (sia in Statistica che negli ambiti delle scienze che utilizzano la Statistica).
Avevo già letto la prima edizione con grande piacere, e questa seconda edizione include interessanti aggiornamenti.
David W. –
I have been reading this book on and off for the past year. Initially the 2nd edition draft was available online as well as the lecture series from 2019. I have a decent statistics background, but felt some gaps in Bayesian so wanted to give it another shot. This book has fantastic applied problems & the author does a great job of breaking down the concepts into an easily digestible way. While I prefer Python, the package that Richard McElreath has put together is very helpful. The lectures of his courses are available online, a great pairing to a great book. This is a must have in my minimalist collection of textbooks!
Mark Saroufim –
Over the years I’ve bought many Bayesian Analysis textbooks, the reason being I knew from ML academics that working with distributions is the “true” way of doing ML instead of just point estimates like in industrial ML.
Before picking up this book I had given up on ever finding a real use-case for Bayesian ML because most of the other often recommended textbooks I owned were happy with long tedious mathematical derivations that wouldn’t even bother explaining why a technique is important or how to implement it.
This book is exceptional in that it gives you the historical context behind how certain techniques were evolved and an excellent intuition for how they work. The book also comes with an R Bayesian Analysis library which also has excellent ports in Julia and Python on Github. In the case where you don’t have gigantic amounts of data and where you’d like to question assumptions that you have about data, this book will teach you a way of how to think about data that is sorely lacking in any sort of Deep Learning text.
All of the algorithms in this book have stood the test of time and will continue to be relevant for the foreseeable future, thankfully this book exists to make these algorithms understandable.
G. F. M. P. –
There is a lecture series on YouTube that is the perfect accompaniment to the book (just search for the author in YT).
The book is basic enough to be understandable to non-mathematicians/ non-statisticians but not so basic that it’s boring/ redundant. The R code examples are great for learning how to use R.
Owen Davis –
Most introductory textbooks on Bayesian inference and statistics are slow and unintuitive and take ages to get to the point. This book is a much-needed exception to the rule. It is written clearly and builds up from the most intuitive fundamentals of probability and statistics. It has wit and charm. The examples and “rethinking” package in R help greatly in illustrating some of the more challenging concepts. Absolutely recommended for anyone wishing to dip a toe into the world of Bayesian inference.
chini mattia –
I wished the book was a bit more dense, with less storytelling and a bit more depth to the arguments that are treated.
I found the DAG chapter on the one hand quite illuminating (it was completely new to me), but on the other hand the explanation was clearer on other sources.
vincent tanoe –
Great book
Cuong Duong –
Lots of positives about this book:
– Accompanying lectures by the author which are available online for free on his YouTube channel
– Author tries to make Bayesian stats as intuitive as possible, and most explanations are by examples and code rather than written math.
– Places heavy emphasis on the use of Bayesian stats for inference rather than predictive modelling (but does explain the importance of good model fit, etc. as well).
– Explains how to set good priors, with examples, which is usually missing in a lot of other instructive material on Bayesian modelling.
Some things to note that might be issues depending on your specific needs:
– Examples are pretty reliant on the rethinking package, instead of pure Stan or rstan. This is a small issue though since there are reference manuals online for how to use those tools (the book is more about teaching the Bayesian way of thinking and causal inference rather than a specific tool).
– There is a focus on the social sciences so there’s little application to ‘bigger data’ domains where distributions are a little different and data size can be an issue for Bayesian inference (e.g. Tech). Book will provide good fundamentals for extending to this kind of domain though.
– Probably not for more intermediate or advanced users of Bayesian stats (e.g. you’ve already built a few models end to end).
rami krispin –
I enjoy reading every page of this book. Explaining statistical concepts in a simple and intuitive manner. highly recommend this book if you have good knowledge of frequency statistics and want to learn Bayesian statistics.
Flavio M. M. Barros –
Antes de falar do livro, só um background a meu respeito, eu sou bacharel em estatística e fiz mestrado na engenharia em aplicações de mineração de dados. Então eu diria que eu tenho uma formação sólida em inferência clássica (ou frequentista) e bastante intimidade com o uso do R, a ferramenta computacional usada nesse livro. Então a minha perspectiva é de alguém com experiência em estatística, mas que está explorando um outro paradigma de inferência, no caso a inferência bayesiana.
O livro vai do completamente básico em estatística até aplicações sofisticadas de métodos bayesianos de análise de dados. O nível matemático exigido é relativamente baixo, e inclusive o autor deixa claro que o livro não demanda um conhecimento profundo de cálculo ou álgebra linear, e o livro faz muito uso do método computacional para o ensino, isto é, são apresentados os conceitos e o leitor tem a oportunidade de implementar os métodos e discutir os resultados ao longo do texto. Mas não se enganem, o livro é direcionado a alguém que conhece e entende o método científico e pretende utilizar a inferência bayesiana para estatística aplicada em nível de pós-graduação. Não é uma introdução superficial apesar de que eu acredito que um graduando, bastante motivado, poderia aproveitar bem esse livro. Mas, por outro lado, é um livro muito gostoso de ler e aprender e o autor apresenta a inferência bayesiana sob uma perspectiva nova na minha opinião. Acho que como introdução ao assunto não tem nenhum livro tão bom quanto esse no mercado.
Alguns destaques que esse livro teve para mim foram:
1) mostrar como a inferência bayesiana é basicamente um processo de contagem;
2) o pacote rethinking do R que é muito útil para usar com o livro mas também para implementar as próprias análises no futuro;
3) os DAGs (direct acyclic graph) e a inferência causal; nunca tinha visto isso e foi um divisor de águas para mim;
4) a discussão sobre entropia e as distribuições de probabilidade;
5) as discussões sobre MCMC, e especialmente sobre o HMC (monte carlo hamiltoniano). Nas aulas online as simulações que mostram a diferença dos algoritmos de Metropolis e do Gibbs para o HMC foram reveladoras para mim;
6) o fato de ter um curso online do livro no Youtube, onde você pode ler o livro e assistir as aulas junto, o que foi uma tremenda experiência para mim;
Fred –
Modern, deep, effective approach to build (causal Bayesian) models.
Taylor Woolf –
This text is excellent for a beginners introduction to Bayesian statistics. I was using JASP for my analyses rather than the authors packages, but it helped a lot with my general understanding of what happening behind the scenes. I look forward to using his R package for my next data project.
Alessandro –
5 stelle nel suo genere, 4 nell’ambito di libri di statistica.
Statistical Rethinking di Richard McElreath e` un libro di statistica con poca matematica e molta parte discorsiva, i 16 capitoli potrebbero anche essere 16 post (molto) lunghi di un blog di statistica Bayesiana.
L’autore cerca di dare piu` una comprensione intuitiva dei vari concetti piuttosto che seguire un approccio basato su definizioni rigorose e concise quindi chi e` abituato ad una trattazione matematica rimarra` probabilmente deluso.
Ma l’intento e` proprio quello di evitare eccessivi o inutili formalismi e concentrarsi su esempi (con codice in R presente nel libro e scaricabile) svolti e ben descritti.
Da qui la mia valutazione di 5 stelle per il genere “poca matematica” mentre se dovessi valutarlo come elemento piu` generale dell’insieme dei libri di statistica darei solo 4 stelle perche’ sono convinto che il giusto formalismo arricchirebbe ulteriormente l’approccio dell’autore (che forse teme di alienarsi un pubblico che secondo certi stereotipi e` refrattario alla matematica).
Lo stile alle volte e`, almeno per me, piacevolmente provocatorio, come puo` leggersi in questi due brani:
In the sciences, there is sometimes a culture of anxiety surrounding statistical inference.
It used to be that researchers couldn’t easily construct and study their own custom models, because they had to rely upon statisticians to properly study the models first.
This led to concerns about unconventional models, concerns about breaking the laws of statistics.
But statistical computing is much more capable now.
Now you can imagine your own generative process, simulate data from it, write the model, and verify that it recovers the true parameter values.
You don’t have to wait for a mathematician to legalize the model you need.
Probability theory is not difficult mathematically. It is just counting.
But it is hard to interpret and apply.
Doing so often seems to require some cleverness, and authors have an incentive to solve problems in clever ways, just to show off.
But we don’t need that cleverness, if we ruthlessly apply conditional probability.
in conclusione: consigliato come primo libro per avvicinarsi alla statistica Bayesiana.
Karl Norden –
This is a good reference book and an excellent tutorial.
Grady Heller –
Very good, accessible, and worth it. While a background in frequentist isn’t required, it is suggested. This book definitely allows you to both learn and apply Bayesian analysis as you would in the “real world,” being more applied than theoretical. Excellent if you’re a scientist or statistician wanting to finally break Bayesian, or just buff up on some R skills.
Ruslan Mamedov –
And it’s a joy to read it!
Ezam –
This book has a good balance between examples, theory, practical application. Even though it includes R code, the GitHub site includes conversion to Python and other programming languages
Dan Jenson –
Clear description, but none of the hyperlinks work. Everything in blue doesn’t link correctly.
isa –
Buen contenido
Amazing Reviewer –
….let this be it. And catch McElreath’s YT lectures. *GOLD*.
He thinks very deeply and explains methodically. Provides exercises.
Amazing value for money.
Samuel Ahuno –
HardCover works for me. The content of this book is immeasurable and I highly recommend it for your statistics & data analysis journey.
Tee –
Imminently readable. Content rich. Highly recommended.
Dr Entropy –
I must confess that I was quite hesitant to pick this book up when I first encountered the strong recommendations of experienced Bayesian practitioners. The ‘cutesy’ chapter titles and topics really threw me off, and as someone who actively uses Bayesian stats I was sure there would not be much for me to learn from this book.
Well, I was quite wrong. This is one of the most enjoyable technical books I have read in a long time and it really helped focus my skills and put the tools of Bayesian stats in perspective. Highly recommended to even experienced data scientists… even when he is covering ground you know well, it gives you a new way to think and communicate it to others.
ACCGTGGTGACA… –
Combine this with Gelman’s more technical Bayesian Data Analysis and you have a master class that will cover >90% of your data science challenges.
TM –
I’ve been a professional statistician for a long time, and I’ve read or tried to read a ton of books. This book covers a a lot of the tools used in day to day practice, provides clearly written useful advice, and has a practical point of view that is both mathematically sound and helps build the reader’s data intuition. One of the best statistics books I’ve ever read.
Clive Fox –
Currently working my way through the lectures on YouTube, then the written chapters and exercises. This seems an ideal way to learn for me. Having been brought up on standard frequentist approaches this is a fantastic way to get into Bayesian approaches which I’ve been wanting to do for some time. It’s quite a long course and book so do plan on spending a couple of months to get most from it i.e. this is not a quick primer.
DL –
This book is fantastic, it clearly explains the fundamentals of Bayesian models and pitfalls with causal inference. Nearly devoid of the excessive mathematic jargon in similar books and instead focuses on helpful examples and R code. Extra points for making the subject appear fun while still going into far more in depth than math heavy alternatives.
Hailey –
Thanks
federicosalvati –
I am a PhD in berlin I have been literally blown away by this book. It does not only teach you statistics. It makes you a better scientist.
Alejandro Alonso Membrilla –
Probably the best book you can read as a newbie in Bayesian modeling and statistical modeling in general. Covers the theory, not getting into much depth, but also the philosophy of science and modeling and the practical implications of each technique. Well written and with many interesting references to foundational writings in each topic.