Bayesian Models for Astrophysical Data
Using R, JAGS, Python, and Stan
- Joseph M. Hilbe, Arizona State University
- Rafael S. de Souza, University of North Carolina, Chapel Hill
- Emille E. O. Ishida, Université Clermont-Auvergne (Université Blaise Pascal), France
This comprehensive guide to Bayesian methods in astronomy enables hands-on work by supplying complete R, JAGS, Python, and Stan code, to use directly or to adapt. It begins by examining the normal model from both frequentist and Bayesian perspectives and then progresses to a full range of Bayesian generalized linear and mixed or hierarchical models, as well as additional types of models such as ABC and INLA. The book provides code that is largely unavailable elsewhere and includes details on interpreting and evaluating Bayesian models. Initial discussions offer models in synthetic form so that readers can easily adapt them to their own data; later the models are applied to real astronomical data. The consistent focus is on hands-on modeling, analysis of data, and interpretations that address scientific questions. A must-have for astronomers, its concrete approach will also be attractive to researchers in the sciences more generally.Read more
- Supplies complete software code in R, JAGS, Python, and Stan for download
- Discusses innovative Bayesian models that advance and improve astronomical research
- Demonstrates and enables hands-on use of models on real astronomical data
- Winner, 2018 PROSE Award for Cosmology and Astronomy
Reviews & endorsements
'This volume is a very welcome addition to the small but growing library of resources for advanced analysis of astronomical data. Astronomers are often confronted with complex constrained regression problems, situations that benefit from computationally intensive Bayesian approaches. The authors provide a unique and sophisticated guide with tutorials in methodology and software implementation. The worked examples are impressive. Many astronomers use Python and will benefit from the less familiar capabilities of R, Stan, and JAGS for Bayesian analysis. I suspect the work will also be useful to scientists in other fields who venture into the world of Bayesian computational statistics.' Eric D. Feigelson, Pennsylvania State University, author of Modern Statistical Methods for AstronomySee more reviews
'Encyclopaedic in scope, a treasure trove of ready code for the hands-on practitioner.' Ben Wandelt, Paris Institute of Astrophysics, Institut Lagrange de Paris, Université Paris-Sorbonne
'This informative book is a valuable resource for astronomers, astrophysicists, and cosmologists at all levels of their career. From students starting out in the field to researchers at the frontiers of data analysis, everyone will find insightful techniques accompanied by helpful examples of code. With this book, Hilbe, de Souza, and Ishida are firmly taking astrostatistics into the twenty-first century.' Roberto Trotta, Imperial College London, author of The Edge of the Sky
'… the focus of the book is not on providing a full understanding of how the distributions arise, but to give guidelines on how to write code for applications, including building multi-level models, and here it succeeds well, and is an excellent resource in conjunction with powerful packages such as STAN and JAGS.' Alan Heavens, The Observatory
25th May 2017 by Bmooers
This book provides examples of a wide range of generalized linear models for continuous and discrete data including count data. The models for count data include the three-parameter NB-P negative binomial model that are not widely available. The approach is mostly Bayesian. The statistical models are used in the freely available JAGS, Stan, and PyMC3 Bayesian data analysis software. These programs use Markov Chain Monte Carlo (MCMC) samplers to estimate the posterior distributions from complex, multilevel statistical models. The JAGS models are run from within the R statistics program. The Stan models are run with Python via PyStan. The last combination is not common because most Stan users run Stan from within R. There are some pure R examples, including several examples that use the first author's COUNT package. The COUNT package was described in two of his recent books on modeling count data. There are also some examples that use the Python package PyMC3. The first author has published almost 20 books about generalized linear models. His books contain a mix of theory and valuable insights from practical experience. The writing style is clear and readily accessible to scientists with a good background in statistics. The book should also appeal to applied statisticians who are building their repertoire of Bayesian data analysis tools and who are looking for fresh statistical models and code for JAGS, Stan, and PyMC3. The book may be ideal for beginning to intermediate users of Bayesian MCMC software. Most examples contain the complete code for both JAGS and PyStan. The data sets are obtained from the book's webpage or the original sources. The code listings are extensive, so there is not room for in-depth explanation on every topic. For example, Hamiltonian Monte Carlo (HMC) is mentioned only once, and the associated No-U-Turn Sampler (NUTS) is not mentioned. NUTS is responsible for Stan's successes with high-dimensional models. Novices need to go elsewhere to fill out their background (e.g., http://mc-stan.org is a rich information resource). The purpose of the book is to promote the application of Bayesian methods with generalized linear models in astrostatistics by providing computer code. The book will also be valuable to a wider audience looking for a cookbook of statistical models and Bayesian data analysis code. The examples are from astrostatistics, but the reader does not have to be an astrophysicist to comprehend the text.
Review was not posted due to profanity×
- Date Published: April 2017
- format: Adobe eBook Reader
- isbn: 9781108216142
- contains: 66 b/w illus. 23 colour illus. 11 tables
- availability: This ISBN is for an eBook version which is distributed on our behalf by a third party.
Table of Contents
3. Frequentist vs Bayesian methods
4. Normal linear models
5. GLM part I - continuous and binomial models
6. GLM part II - count models
7. GLM part III - zero-inflated and hurdle models
8. Hierarchical GLMMs
9. Model selection
10. Astronomical applications
11. The future of astrostatistics
Appendix A. Bayesian modeling using INLA
Find resources associated with this titleYour search for '' returned .
Type Name Unlocked * Format Size
This title is supported by one or more locked resources. Access to locked resources is granted exclusively by Cambridge University Press to lecturers whose faculty status has been verified. To gain access to locked resources, lecturers should sign in to or register for a Cambridge user account.
Please use locked resources responsibly and exercise your professional discretion when choosing how you share these materials with your students. Other lecturers may wish to use locked resources for assessment purposes and their usefulness is undermined when the source files (for example, solution manuals or test banks) are shared online or via social networks.
Supplementary resources are subject to copyright. Lecturers are permitted to view, print or download these resources for use in their teaching, but may not change them or use them for commercial gain.
If you are having problems accessing these resources please contact email@example.com.
Sorry, this resource is locked
Please register or sign in to request access. If you are having problems accessing these resources please email firstname.lastname@example.orgRegister Sign in
You are now leaving the Cambridge University Press website. Your eBook purchase and download will be completed by our partner www.ebooks.com. Please see the permission section of the www.ebooks.com catalogue page for details of the print & copy limits on our eBooks.Continue ×