Autumn 2017

We meet on Tuesdays from 1:30-2:30 in Padelford C-14A, which is on the lower level (this is the CSSS Conference Room). We will focus on learning about a set of three computer software packages commonly used for inference in Bayesian spatial statistics (INLA, spBayes, and Stan). The first six weeks of the quarter will be devoted to learning about these packages, with two weeks for each package. The first week of a pair will cover the main paper and relevant theory behind the package, while the second week will focus on using the package itself. In the last three weeks of the quarter we plan to focus on making comparisons in performance and usage among the packages.

Schedule

The following is the current schedule of topics and presenters. We will continue the sign up process over the first few meetings of the quarter.

Date

Speaker

Topic
October 10 Johnny Paige Introduction to ideas, methods, and math behind INLA -- will focus on Rue et al 2009. Here are the slides from the presentation.
October 17 Hannah Director Intro to R-INLA package and how INLA concepts are implemented
October 24 Jim Faulkner Theory behind spBayes. Here are the slides
October 31 Max Schneider spBayes demo -- Meeting Cancelled
November 7 Sahar Zangenah Intro to Stan [slides]
November 14 Mingwei Tang Theory of HMC and NUTS. Reading: Neal (2011) and Hoffman and Gelman (2014)Here are the slides from the meeting.
November 21 Peter Gao and Serge Aleshin-Guendel Stan worked examples: Example 1 and Example 2 and slides from meeting
November 28 John Best Comparison of packages I -- paper and code from paper, and SLIDES from meeting
December 5 Max Schneider Comparison of packages II

Resources Related to Packages

These are links to the main websites and related papers of the packages we will cover this quarter.

INLA

  • R-INLA project website. This page has a wealth of information, including many tutorials. Note that there is an R interface to INLA but it is available for download only on the R-INLA project website and not on R CRAN.
  • Main INLA paper PDF: Rue, H., S. Martino, and N. Chopin. 2009. Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations. Journal of the Royal Statistical Society, Series B 71(2):319-392.
  • Paper on spatial analysis with R-INLA -- Page with PDF and supplementary code: Lindgren, F., and H. Rue. 2015. Bayesian spatial modelling with R-INLA. Journal of Statistical Software 63(19):1-25.
  • Paper with extensions spatial analysis with R-INLA -- Page with PDF and supplementary code: Bivand, R., V. Gomez-Rubio, and H. Rue. 2015. Spatial data analysis with R-INLA with some extensions. Journal of Statistical Software 63(20):1-31.
  • Original paper describing SPDE approach: Lindgren, F., and H. Rue. 2011. An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach. Journal of the Royal Statistical Society, Series B 73(4):423-498. PDF
  • Book on spatial stats using R-INLA: Blangiardo, M., and M., Cameletti. 2015. Spatial and spatio-temporal Bayesian models with R-INLA. John Wiley and Sons, Ltd. E-Book with UW access

spBayes

  • R CRAN page for spBayes.
  • Original software paper on spBayes -- Page with PDF and supplementary code: Finley, A., S. Banerjee, and B. Carlin. 2007. spBayes: an R package for univariate and multivariate point-referenced spatial models. Journal of Statistical Software 19(4):1-24.
  • Updated software paper on newer version of spBayes -- Page with PDF and supplementary code: Finley, A., S. Banerjee, and A. Gelfand. 2015. spBayes for large univariate and multivariate point-referenced spatio-temporal data models. Journal of Statistical Software 63(13):1-28.
  • Slides on some basics of spBayes PDF.
  • Book using spBayes: Banerjee, S., B. P. Carlin, and A. E. Gelfand. 2015. Hierarchical modeling and analysis for spatial data. . Second edition. CRC Press, Taylor & Francis Group.

Stan

  • Stan project website. Check out the Rstan page and also the main documentation page which has tutorials and case studies. Also make sure you check out the Stan Users Forum.
  • R CRAN page for rstan.
  • Software paper on stan -- Page with PDF and supplementary code: Carpenter, B., A. Gelman, M. Hoffman, D. Lee, B. Goodrich, M. Betancourt, M. Brubaker, J. Guo, P. Li, and A. Ridell. 2017. Stan: a probabilistic programming language. Journal of Statistical Software 76(1):1-32.
  • Paper on HMC -- PDF: Neal, R. 2011. MCMC using Hamiltonian dynamics. Chapter 5, Handbook of Markov Chain Monte Carlo. Chapman and Hall.
  • Conceptual intro to HMC -- PDF: Betancourt, M. 2017. A conceptural introduction to Hamiltonian Monte Carlo. arXiv:1701.02434v1 [stat.ME]
  • Paper on adaptive HMC algorithm used in stan -- PDF: Hoffman, M., and A. Gelman. 2014. The no-u-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo. Journal of Machine Learning Research 75:1593-1623.
  • Paper on automatic differntiation used in stan -- PDF: Carpenter, B. M. Hoffman, M. Brubaker, D. Lee, P. Li, and M. Betancourt. 2015. The stan math library: reverse-mode automatic differentiation in C++. arXiv:1509.07164 [cs.MS].
  • Paper with example code for fitting Gaussian processes in stan -- PDF
  • Spatial example with ICAR model: [link].
  • Gaussian process example: [link].
  • Example of Gaussian predictive process models in stan: [link].
  • Example of geostatistical modelling with stan: [link].

Resources for package comparisons

  • Heaton et al. 2017. Methods for analyszing large spatial data: areview and comparison. arXiv:1710.05013v1 [stat.ME] -- PDF
  • Code related to Heaton et al. 2017: -- LINK

Additional Resources

  • Cressie, N., and C. K. Wikle. 2011. Statistics for spatial-temporal data. John Wiley & Sons
  • Banerjee, S., B. P. Carlin, and A. E. Gelfand. 2015. Hierarchical modeling and analysis for spatial data. . Second edition. CRC Press, Taylor & Francis Group.
  • Gelfand, A. E., P. Diggle, P. Guttorp, and M. Fuentes (Eds.). 2010. Handbook of spatial statistics. CRC Press.
  • Rue, H., and L. Held. 2005. Gaussian Markov random fields: theory and applications. Chapman & Hall/CRC.

Assorted Space-Time Links