On demand: Introduction to GAM and GAMM using frequentist tools

 

Generalised Additive Modelling in R

 

Part I — Introduction to GAM and GAMM

Part II — Spatial and Spatio-Temporal Modelling (separate registration required)

 

 

Part I — Introduction to GAM and GAMM

This online course provides a practical and applied introduction to generalised additive models (GAMs) and generalised additive mixed-effects models (GAMMs) in R. The course consists of five modules representing approximately 40 hours of work and is designed for participants working with ecological, environmental, biological, aquaculture, fisheries, or other applied datasets.

The course combines short theory presentations with a large number of hands-on exercises using real datasets. All exercises include annotated R code, downloadable datasets, and on-demand video discussions of the statistical analyses and R solution files.

All video material is available on-demand and can be watched online, as often as you want, at any time of the day, within a 12-month access period.

Participants also receive:

  • Access to an online discussion board for questions and interaction with instructors and participants.
  • Downloadable course material and annotated R scripts,
  • A 1-hour individual face-to-face video consultation with one or both instructors to discuss their own data or statistical questions. The 1-hour consultation must be used within 12 months after receiving access to the course website and needs to be consumed in a single session at a mutually convenient time.

Course content

The course starts with a short revision of data exploration and multiple linear regression. We then introduce generalised additive models for modelling non-linear relationships using the mgcv package in R. Participants will learn how to:

  • fit and interpret GAMs and GAMMs,
  • model non-linear relationships,
  • include random effects and hierarchical structures,
  • apply GAMs to count, proportional, binary, and continuous data,
  • perform model validation using DHARMa,
  • and interpret smoother behaviour and model output.

The course covers:

  • Gaussian GAM and GAMM,
  • Poisson GAM and GAMM,
  • negative binomial GAM and GAMM,
  • Bernoulli and binomial GAM and GAMM,
  • beta GAM,
  • gamma GAM,
  • Tweedie GAM,
  • hierarchical GAMM,
  • univariate and two-dimensional smoothers, and
  • interactions involving smoothers.

The course is highly applied in nature and focuses on practical modelling skills rather than mathematical derivations.

 

Part II — Spatial and Spatio-Temporal GAMM (advanced add-on module)

 

Part II is an advanced continuation of the GAM and GAMM course series and focuses on the analysis of spatial and spatio-temporal data using generalised additive models in R.

Through a combination of theory presentations and hands-on exercises using real datasets, participants will learn how to model spatial structure, temporal trends, and spatial-temporal interactions. The course covers:

    • spatial smoothers,
    • two-dimensional smoothers,
    • spatial GAM and GAMM,
    • spatial-temporal modelling,
    • soap-film smoothers for study areas with barriers,
    • variograms and spatial diagnostics,
    • and the analysis of count, proportional, binary, and continuous spatial data.

    The course is highly applied in nature and focuses on practical modelling skills using modern GAM and GAMM tools in R, primarily through the mgcv package and associated spatial-analysis packages.