Urban pollination networks with stochastic block models
Material of my 3-hour course at the school Biodiversité urbaine : enjeux et modélisation (Institut Henri Poincaré, Paris, 25–27 November 2026): analysis of urban pollination networks with stochastic block models (SBM, colSBM) and network comparison.
Running example: Baldock et al. (2019), Nature Ecology & Evolution 3: 363–373. Plant–pollinator networks in 4 UK cities × 9 land uses.
Slides
R companion
- R companion of the slides: the R code that produces the analyses of the slides, step by step, with its results.
- Code and data (zip): the Quarto source of the companion (
companion_slides_UK.qmd), the data of Baldock et al. (2019) with the script that builds the networks from them, and the estimation scripts (scripts/). - Estimation results (zip): the fitted models (
.rds). Unzip it inside the code folder, so that the files are inresults_analysis/: the companion then reads the fits instead of re-estimating them (some colSBM fits take about 30 min each).
Render the companion from the code folder with quarto render companion_slides_UK.qmd.
R packages: dplyr, tidyr, ggplot2, patchwork, sbm, colSBM (≥ 0.4.11, GitHub GrossSBM/colSBM), bipartite, vegan.