Computational Statistics in Python
Preface
The book Computational Statistics in Python is a summary and adaptation of the book Computational Statistics in Java, published by Editora UFLA. The objective of this book is, first, to improve the author’s own knowledge of Quarto and, second, to consolidate and advance the learning of Python.
We will initially adapt the course notes Computational Statistics in R to accommodate the Python code. Subsequently, we will expand the content of this first edition of the book.
To learn more about Quarto Books, visit https://quarto.org/docs/books.
In these lecture notes, our intention was to address the topic of computational statistics, which is so crucial to the scientific community and especially to graduate students in statistics. We can state without hesitation that computational statistics has become, and is today, one of the main subfields of statistics. Furthermore, knowledge from this area can be—and frequently is—applied to other fields of statistics, engineering, and physics. Bayesian inference is a prime example where a computational approach is typically required. When planning these lecture notes, we had many doubts about what topics to cover and how to approach each selection. Ultimately, we chose to write something that enables the reader to go beyond a simple recipe book while avoiding getting bogged down in a tangle of proofs. On the other hand, we sought to present models and methods in a broad, non-restrictive manner.
Another motivation that guided and encouraged us to develop this project was our personal research experience in computational statistics. We also cannot deny that we undertook this project for personal benefit, coming into direct contact with the vast body of literature published in this branch of statistics. However, we do not intend to cover every subject, nor do we aim to exhaust all possibilities for any given topic. On the contrary, we hope these notes serve as an introduction to computational statistics and inspire undergraduate statistics students to delve even deeper into this field.
These notes are based on a book we wrote on computational statistics using Java. Adapting some of the routines implemented in that book to Python was a very enjoyable and revealing task. We took this opportunity to uncover some of the many secrets of this powerful tool and discover its vast potential—as well as its limitations. In this first version, we do not expect perfection; we are fully aware that flaws may exist and rely on reader feedback to help correct them. We are launching this initial, unrevised, and unexpanded version using Quarto and transitioning from R to Python. These notes will be updated continuously online. We hope this manuscript contributes to the professional growth of students in our graduate program in Statistics and Agricultural Experimentation, as well as students from other programs at UFLA or other institutions. If so, our goal will have been achieved.