I wrote this book to introduce the main concepts and practices of Bayesian statistics to those who are already familiar with Python and the Python data stack, but not very familiar with statistical analysis. Having read the previous eight chapters, you should have a reasonable practical understanding of many of the main topics of Bayesian statistics. Although you will not be an expert-Bayesian-ninja-hacker (whatever that could be), you should be able to create your own probabilistic models to solve your own data analysis problems. If you are really into Bayesian statistics, this book will not be enough – probably no single book will be enough. To become more fluent in Bayesian statistics, you will need practice, time, patience, enthusiasm, and more practice, and you will need to revisit ideas and concepts from a different perspective.

In the repository (https://github.com/aloctavodia/BAP), you will find examples complementing those that are discussed in this book. These are examples that did not fit this book, either due to space or time. In fact, at the time of writing this book, there are no extra examples yet, but I will add examples there from time to time. To gather extra material, you should also check the PyMC3 documentation at https://docs.pymc.io, especially the examples section, which is full of many examples of models that were covered in this book and many others that were not. As you already know, ArviZ is a really new library, but we are already writing an educational resource about exploratory analysis of Bayesian models. We hope this will be a useful reference, especially for newcomers to Bayesian modeling (https://github.com/arviz-devs/arviz_resources).

If you find mistakes in this book, either text or code, you can file an issue at https://github.com/aloctavodia/BAP. If you have general questions about Bayesian statistics, especially those related to PyMC3 or ArviZ, you can ask questions at https://discourse.pymc.io/.

In the next few paragraphs, I list some material that has definitely influenced my Bayesian way of thinking. This list is by no means exhaustive. I am confident that you will also find at least part of this material very useful and inspiring.

If you want to keep learning about Bayesian statistics in general, check out this list:

• I strongly recommend that you read Statistical Rethinking by Richard McElreath. This is a superb introductory book about Bayesian Analysis. The problem :-) is that the examples are in R/Stan. Hence, a group of volunteers have ported the examples in this book to Python/PyMC3. Check out the GitHub repository for more information: https://github.com/pymc-devs/resources/tree/master/Rethinking.
• Another book that's been ported to PyMC3 is Doing Bayesian Data Analysis by John K. Kruschke (also known as the puppy book). This is another nice introductory book about Bayesian analysis. Most of the examples from the first edition of this  book have been ported to Python/PyMC3 in the following GitHub repository: https://github.com/aloctavodia/Doing_bayesian_data_analysis. You can find the second edition here: https://github.com/JWarmenhoven/DBDA-python. Unlike Statistical Rethinking, the puppy book is more focused on how to carry the Bayesian analog of many commonly frequentist statistical analysis. Depending on what you want, this can be a pro or a con of this book.
• Allen B. Downey has many great books, and Think Bayes (http://greenteapress.com/wp/think-bayes/) is not an exception. In this book, you will find several interesting examples and scenarios that will certainly challenge you and help you grasp the Bayesian approach to problem solving. This book does not use PyMC3, but a Python library that's constructed around Think Bayes. The second edition, which is still not written at the moment of publishing this book, will use PyMC3 and probably even ArviZ. You can check the repository for this second edition: https://github.com/AllenDowney/ThinkBayes2
• Another (optionally free/paid resource) is Probabilistic Programming and Bayesian Methods for Hackers, by Cameron Davidson-Pilon and several contributors. This book/notebook was originally written using PyMC2 and has now been ported to PyMC3: https://github.com/quantopian/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers.
• One book that is generally refereed as The Bayesian book is Bayesian Data Analysis by Andrew Gelman and others. This is definitely a great book, although it is not really an introductory book and probably works better as a reference book and not a textbook. If you are not familiar with statistics (Bayesian or otherwise), I recommend that you first pick Statistical Rethinking by Richard McElreath and then try Bayesian Data Analysis. You may also want to check the book Data Analysis Using Regression and Multilevel/Hierarchical Models, by Andrew Gelman and Jennifer Hill.
  

If you want to keep learning about Gaussian processes, check out the following book:

• Gaussian Processes for Machine Learning, by Carl Edward Rasmussen and Christopher K. I. Williams, is the book for Gaussian processes. It was awarded with the 2009 DeGroot Prize of the International Society for Bayesian Analysis, and the only downside is that we all want a new edition!

A couple of machine learning books with a Bayesian twist are as follows:

• Machine Learning: A probabilistic Probabilistic Perspective, by Kevin P. Murphy. This is a great book that tries to explain how many methods and models work by using a probabilistic approach. You may find this book a little dry or very concise and to the point, depending on your mathematical inclinations. Either way, this book is full of examples and was written with a very practical aim. Kevin Murphy has taken examples and ideas from many other sources and thus this book is a great summary of many other great resources. The first time I heard about deep learning was from this book, way before it became the new cool kid on the block.
  
• Pattern Recognition and Machine Learning, by Christopher Bishop, is a classical book in machine learning and has considerable overlap with Machine Learning: A probabilistic Probabilistic Perspective, although probably with a little bit more of a Bayesian perspective. It is also maybe a little bit easier to read as a textbook than Murphy's, which is more of a reference book.

As a child, I dreamed of flying cars, clean unlimited energy, vacations on Mars or the Moon, a global government pursuing the well being of the entire human race.. yeah, I know... I used to be a dreamer! For many reasons, we have none of that. Instead, we have something that was completely unimagined, at least for me, just a couple of decades ago: the democratization of very powerful computer methods. One of the side effects of the computer revolution is that any person with a modest understanding of a programming language like Python now has access to a plethora of computational methods for data analysis, simulations, and other complex tasks. I think this is super-great, but also an invitation to be extra careful about these methods. The way I learned about statistics as an undergrad and how I had to memorize how to use canned methods was frustrating, useless, and completely unrelated to all of these changes. At a very personal level, this book is perhaps a response to that frustrating experience.

I tried to write a statistical book with an emphasis on a modeling approach and a judicious context-dependent analysis. I am unsure whether I had really succeeded in this front. One reason for this is probably that I still need to learn more about this (maybe we as a community need to learn more about this). Another reason is that a proper statistical analysis should be guided by the domain-knowledge and context, and providing context is generally difficult in a book with a very broad target audience. Nevertheless, I hope that I have provided a sane, skeptical perspective regarding statistical models, some useful examples, and enough momentum for you to keep learning.