Robbie M. Parks and Cascade Tuholske. With thanks to Marianthi-Anna Kioumourtzoglou and Garyfallos Konstantinoudis for previous work together informing this material.
GitHub repo through which we're developing and sharing materials for the Introduction to methods in environmental epidemiology Workshop, taking place in person during July 21st-23rd 2025.
The workshop will be a series of lectures and interactive supervised lab sessions. We hope it's informative and fun!
We will be using Posit (RStudio) Cloud, which assumes knowledge of R and RStudio. We will also ask you to pull the final versions of the GitHub repo to your Cloud account. The basics of doing this are in a previously-created guide found via another guide repo.
This Bayesian component of this workshop is largely written in NIMBLE.
Cascade's section of the workshop will be in Python.
Below is the set of labs to follow throughout the three days:
| Time | Activity |
|---|---|
| 9:00 - 10:00 | Introduction to Methods in Environmental Epidemiology 1 (Lecture) - Theories and assumptions, Data structuring |
| 10:00 - 11:00 | Introduction to Methods in Environmental Epidemiology 1 (Hands-on Lab) - Theories and assumptions, Data structuring |
| 11:00 - 11:15 | Break |
| 11:15 - 12:15 | Introduction to Methods in Environmental Epidemiology 2 (Lecture) - Model likelihood structures (Normal, Poisson, Bernoulli, Binomial) |
| 12:15 - 13:15 | Lunch |
| 13:15 - 14:15 | Introduction to Methods in Environmental Epidemiology 2 (Hands-on Lab) - Model likelihood structures (Normal, Poisson, Bernoulli, Binomial) |
| 14:15 - 15:15 | Introduction to Methods in Environmental Epidemiology 3 (Lecture) - Quantile regression |
| 15:15 - 15:30 | Break |
| 15:30 - 16:30 | Introduction to Methods in Environmental Epidemiology 3 (Hands-on Lab) - Quantile regression |
| 16:30 - 17:00 | Questions and Wrap-up |
| Time | Activity |
|---|---|
| 9:00 - 10:00 | Modeling Non-linear and Delayed Time Series Effects 1 (Lecture) - Non-linear exposure-response curves |
| 10:00 - 11:00 | Modeling Non-linear and Delayed Time Series Effects 1 (Hands-on Lab) - Non-linear exposure-response curves |
| 11:00 - 11:15 | Break |
| 11:15 - 12:15 | Modeling Non-linear and Delayed Time Series Effects 2 (Lecture) - Case crossover design , Time series design |
| 12:15 - 13:15 | Lunch |
| 13:15 - 14:15 | Modeling Non-linear and Delayed Time Series Effects 2 (Hands-on Lab) - Case crossover design , Time series design |
| 14:15 - 15:15 | Modeling Non-linear and Delayed Time Series Effects 3 (Lecture) - Distributed lag non-linear models (DLNM) , Case crossover with hybrid DLNM |
| 15:15 - 15:30 | Break |
| 15:30 - 16:30 | Modeling Non-linear and Delayed Time Series Effects 3 (Hands-on Lab) - Distributed lag non-linear models (DLNM). Case crossover with hybrid DLNM |
| 16:30 - 17:00 | Questions and Wrap-up |
| Time | Activity |
|---|---|
| 9:00 - 10:00 | Introduction to Bayesian Methods 1 (Lecture) - Brief history of Bayesian methods, Overview of Bayesian methods |
| 10:00 - 11:00 | Introduction to Bayesian Methods 1 (Hands-on Lab) - Brief history of Bayesian methods, Overview of Bayesian methods |
| 11:00 - 11:15 | Break |
| 11:15 - 12:15 | Introduction to Bayesian Methods 2 (Lecture) - Bayesian regression, Hierarchical modelling |
| 12:15 - 13:15 | Lunch |
| 13:15 - 14:15 | Introduction to Bayesian Methods 2 (Hands-on Lab) - Bayesian regression, Hierarchical modelling |
| 14:15 - 15:15 | Introduction to Heat Exposures 1 (Cascade) |
| 15:15 - 15:30 | Break |
| 15:30 - 16:30 | Introduction to Heat Exposures 2 (Cascade) |
| 16:30 - 17:00 | Questions and Wrap-up |
Run pre-commit install to install the hooks. You now won't be able to commit until you pass the hooks. These (among other things) automatically format files and prevent us from committing ugly code. For more details, see the main docs and the R docs.
renv maintains consistency between users' R environments. Run renv::restore() and the environment will be downloaded into the repository based on the renv.lock file. If you want to add a packages to the lockfile, install the package and then run renv::snapshot(). For more details, see the docs.
Quarto is pretty cool. I won't bore you, but have a look at the docs. Here, we're using it for presentations. It's designed by the folks at RStudio, so you R folk will be happy. Make a .qmd file and run quarto render *.qmd to generate the html, which you can open in browser. We can get fancy and import our own css to have a consistent theme for out presentations.