Learning

Notebooks made with quarto + R code. These are personal study material: I wrote them myself to learn and practice, sometimes adapting or copy-pasting pieces from other people’s work along the way. I share them here as drafts, in case they’re useful to scientific collaborators or students too.

Open Science

Registered Reports: what are they? — I explain the two-stage Registered Report format, how it differs from a preregistration, and where to find real Stage 1 and Stage 2 examples in ecology.

Ecological Modelling

Empirical Dynamic Modelling: an introduction — I introduce state-space reconstruction, Simplex projection, and S-map using rEDM, and show you how to tell a chaotic population from a noisy one.
Convergent Cross Mapping — I use rEDM to test causal links between time series, on both a synthetic system and a classic predator-prey dataset, and address whether CCM needs to control for confounders the way regression does.
Ecopath with Ecosim, and its R counterpart Rpath — I introduce the EwE mass-balance framework, then rebuild my published Rocas Atoll food-web model in Rpath and run it forward under warming scenarios — including where the model stops being believable.
Gaussian Processes — I introduce Gaussian Process regression and the GP-EDM framework, working through the maths and then a few ecological toy examples.

Statistics

Priors — I visualize a handful of commonly used priors (Gamma, Exponential, Half-Cauchy, and more) so I can build some intuition about what they imply before I use them.
Measurement Error — I use brms to simulate how measurement error in a predictor and/or an outcome distorts a simple regression’s coefficients.
Sample Size Planning — I simulate an ecological dataset at increasing sample sizes to show you, concretely, how the number of observations you collect shapes the power and precision of a simple linear brms model.