Workshop Materials

Main R markdown document

Rmd:

HTML:

Data Sets

Put these in a folder called data:

Other Stan Resources

Additional resources will be posted here.

Teaching Resources

Dates and Location

Calvin College will be hosting a Stan (+R) workshop June 10-12. Sessions will begin each day at 9 am and conclude at 4 pm. Lunch and mid-session refreshments will be provided.

Calvin College is located in Grand Rapids, MI. Campus Map

There should be ample parking in lots 4 or 5 (the closest to North Hall). Parking Lot Map

Before You Arrive

To get the most out of the workshop, come with a laptop ready to use RStudio and Stan. (Let the organizers know ASAP if you do not have a laptop you can bring and we will see if we can find something for you to borrow.) Here’s is what should be on your laptop.

  • A recent version of RStudio (1.2.x or later)
  • A recent version of R
  • A recent version of Stan (version 2.18.0 or above).

  • The following R packages

    packages <- c('bayesplot', 'loo', 'shinystan', 'rstanarm', 
              'brms', 'lubridate',  'rmarkdown', 'tidyverse')
install.packages(packages)

Target Audience

The target audience is faculty and other researchers, either teachers or practitioners of applied statistics.

Registration

Limited space is available for this workshop. Registration via EventBrite is available at https://calvin-stan-workshop.eventbrite.com.

Update: We are sold out, but we are maintaining a waiting list.

Workshop Description

This three-day workshop will provide an introduction to using Stan for Bayesian data analysis.

What is Stan?

Stan® is a state-of-the-art platform for statistical modeling and high-performance statistical computation. Users specify log density functions in Stan’s probabilistic programming language and get:

  • full Bayesian statistical inference with MCMC sampling (NUTS, HMC)

  • approximate Bayesian inference with variational inference (ADVI)

  • penalized maximum likelihood estimation with optimization (L-BFGS)

Workshop sessions will include information on designing models, choosing priors, describing models in the Stan language, detecting and addressing potential problems with the HMC sampler, and interpreting results. Our primary interface to Stan will be via R and RStudio, but participants should be able to transfer skills to other interfaces, including python. Ample hands-on time will be provided with guided activities to give participants experience using Stan. No prior experience with Stan is required.

Presenters

Our lead presenter will be Jonah Sol Gabry. Jonah is a member of the Stan core development team and a researcher in statistics working with Andrew Gelman on methods and software for Bayesian data analysis. He is co-author of the rstan and rstanarm R packages, which provide interfaces to Stan, as well as author and maintainer of the shinystan and bayesplot packages for model visualization, and the loo R package for model comparison. Jonah is also affiliated with the Columbia Population Research Center, where he advises on statistical issues related to collection and analysis of survey data.

Joining Jonah will be Vianey Leos Barajas. Vianey is a statistician who does research in the area of statistical ecology, with a primary focus on time series modeling of animal movement data. She currently uses Stan to model shark and elasmobranch data in collaboration with the MigraMar group.

Lodging and Transportation

Participants from out of town can choose to stay at the Prince Conference Center ($129 + tax per night) on Calvin’s campus (mention the Stan Workshop when reserving your room) or at one of many nearby hotels.

The Gerald R. Ford International Airport (GRR) is located approximately 6 miles from campus. The Prince Conference Center provides shuttle service to/from the airport for guests who request this in advance.

Sponsors

This workshop is made possible by the Vos Endowment for Excellence in Mathematics and Statistics of Calvin College and through support of the Department of Statistics at Grand Valley State University.