In no particular order:
- Generalized Additive Models in R: A Free Interactive Course, by me. A friendly introduction requiring only basic knowledge of R and linear regression. 4-5 hours of slides and interactive exercises. This was formerly on a commercial platform but is now open source.
- Slides from my talk at the New York R User's Group (this repo). This is a high-level overview of things that GAMs and mgcv can do. Video (~80 minutes) here: https://www.youtube.com/watch?v=q4_t8jXcQgc
- Materials from our 2017 Ecological Society of America workshop on GAMs. These are designed for the interactive workshop but may still be useful. Target audience is graduate students with a little more statistical training. GLM knowledge a prerequisite. More material on inference, theory and more exotic distributions. See the references page in particular
- The 2018 materials are almost exactly the same, but we are tracking issues in that repo for future improvement!
- Gavin Simpson's ~3 hour YouTube introduction to GAMs covers much of the same material as my course in one long lecture, with different examples and some updates as of summer 2020.
- The essential GAMs reference is Simon Wood's Generalized Additive Models in R.
- Recently reviewed by Virgilio Gomez-Rubio in the Journal of Statistical Software
- An online book by Michael Clark gives an a very nice short introduction to both GAM theory and use in R.
- Gavin Simpsons's excellent, GAM-centric blog where he tries out new and little-used GAM formulations.
- StackOverflow and Cross Validated tags for
mgcv
- Home to Gavin's amazing Cross Validated answer on spatiotemporal modelling with GAMs
- A post by Kim Larsen's GAMs on the Stitchfix Blog which explains GAMs and compares them to other methods for classification.
- The gratia package by Gavin Simpson for using mgcv with ggplot2 and other useful and tidy helper functions, such as calculating spline derivatives and simulating from model posteriors. Here's a blog post introducing it.
- Hierarchical Generalized Additive Models: an introduction with mgcv A paper by Eric J. Pedersen, David L. Miller, Gavin Simpson, and Noam Ross on fitting gams with heirarchical/mixed structures. GitHub repo here.
- Modelling palaeoecological time series using generalized additive models, a paper by Gavin L. Simpson
- Bayesian views of generalized additive modelling, a brief but useful write-up on Bayesian approaches and interpretations of GAMs by Dave Miller.
- Generalised additive mixed models for dynamic analysis in linguistics: a practical introduction by Márton Sóskuthy
- Simplified Integrated Nested Laplace Approximation by Simon N. Wood, details the
ginla()
function added in mgcv 1.8-27.