We compare the methClust model fit on the sites by bird species presence absence matrix with the bird species by sites presence absence matrix.

Plotting function of clusters

latlong <- get(load("../data/LatLongCells_frame.rda"))
world_map <- map_data("world")
world_map <- world_map[world_map$region != "Antarctica",] # intercourse antarctica

p <- ggplot() + coord_fixed() +
  xlab("") + ylab("")

#Add map to base plot
base_world_messy <- p + geom_map(data=world_map, map = world_map, aes(group=group, map_id=region), colour="white", fill="white", size=0.05, alpha=1/4) 

cleanup <- 
  theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), 
        panel.background = element_rect(fill = 'white', colour = 'white'), 
        axis.line = element_line(colour = "white"), legend.position="none",
        axis.ticks=element_blank(), axis.text.x=element_blank(),
        axis.text.y=element_blank())

base_world <- base_world_messy + cleanup
PlotAssemblageIdx <- function(dat){
  dat <- cbind.data.frame(latlong, dat)
colnames(dat) <- c("Latitude", "Longitude", "Value")
map_data_coloured <- 
  base_world +
  geom_point(data=dat, 
             aes(x=Latitude, y=Longitude, colour=Value), size=0.5) +
  scale_colour_gradient(low = "white", high = "black") 

map_data_coloured
}

K = 2

(cells -> birds)

K2_global_topics <- get(load(file="../output/birds_meth_cluster_2.rda"))

topics_omega <- K2_global_topics$omega
PlotAssemblageIdx(topics_omega[,1])

PlotAssemblageIdx(topics_omega[,2])

(birds -> cells)

Results for K = 2:

K = 3

(cells -> birds)

(birds -> cells)

Results for K = 3:

K = 5

(cells -> birds)

(birds -> cells)

Results for K = 5:

K = 10

(cells -> birds)

(birds -> cells)

Results for K = 10:

K = 11

(cells -> birds)

(birds -> cells)

This R Markdown site was created with workflowr