fdschneider / caspr

Cellular Automata for Spatial Pressure in R
MIT License
4 stars 0 forks source link

(for reference) Models bifurcation diagrams #39

Open alexgenin opened 8 years ago

alexgenin commented 8 years ago

In order to generate the right datasets, here are the bifurcation diagrams for the 4 models we have so far (some of them work flawlessly while some others might not) :

Forest gap model (alpha: 0.2, d: 0.01, delta: 0-1)

ggplot(results[['forestgap']]) + 
  geom_point(aes(delta, mean_cover_.)) + 
  ylab('Forest global cover') 

unnamed-chunk-2-1

Grazing model (del: 0.9, b: 0-1, c_: 0.2, m0: 0.05, g: 0.2, r: 0.01, f: 0.9, d: 0.1, p: 1)

ggplot(results[['grazing']]) + 
  geom_point(aes(b, mean_cover_.)) + 
  ylab('Veg global cover')

unnamed-chunk-3-1

Mussel bed model (r: 0.7, d: 0.1, delta: 0-0.75)

ggplot(results[['musselbed']]) + 
  geom_point(aes(delta, mean_cover_.)) + 
  ylab('Mussel global cover')

unnamed-chunk-4-1

Predator-prey model (betaf: 1/3, betas: 0.1, delta: 0-0.1)

ggplot(results[['predprey']]) + 
  geom_point(aes(delta, mean_cover_f)) + 
  ylab('Prey global cover')

unnamed-chunk-5-1

grazing and forestgap models seem to work flawlessly. I'm a bit more hesitant about the other two (and there is no consensus implementation across all the articles/code snippets available), so any previous experience with those model is appreciated.