CODEWITHSUNDEEP
rplotvegan
MyNameisTK
MyNameisTK

Reputation: 229

accumulation curve in R

I have data of species at 4 sites over several months. I have successfully created accumulation graphs using package vegan in R but I would like to plot all 4 sites on one graph.

At first I had a data sheet with all sites and months but the when I plotted specaccum the result was a curve of accumulation of all data regardless of site.

Because of this I split each site into a separate data sheet which I loaded into R. In each data sheet the first row is species names and each additional row below that is a month.

For example I loaded the data of one of my sites "FMR". Then I did the following:

FMR <-specaccum(FMRJ2F, "random")
plot(FMR)

I did the same for my other sites, PP, DUX, PM. How can i put all 4 lines on one plot?

Upvotes: 7

Views: 15385

Answers (2)

jlhoward
jlhoward

Reputation: 59345

You can just use the add=TRUE argument in plot.specaccum().

library(vegan)
data(BCI) 
l <- lapply(c(1,21,41,61,81), \(i) specaccum(BCI[, seq(i,i+19)], method="random"))
plot(l[[1]])
for (i in 2:5) plot(l[[i]], add=TRUE, col=i)

This code snippet just loads the built-in BSI dataset from {vegan}, and creates a list l of 5 specaccum objects by running specaccum() on a subset of the columns in BCI. You don't need to do this since you already have the specaccum objects.

Then, we create the first plot, and add, with each iteration of the for-loop, a new curve with add=TRUE.

Upvotes: 8

jbaums
jbaums

Reputation: 27388

Ok, so @jlhoward's solution is of course much simpler, and more sensible. But, since I hadn't thought of the obvious, and coded this up, I figured I may as well share it. It might be useful for related questions where the function at hand doesn't accept add.

Load library and some example data:

library(vegan)
data(BCI)
sp1 <- specaccum(BCI, 'random')

# random modification to BCI data to create data for a second curve
BCI2 <- as.matrix(BCI)
BCI2[sample(prod(dim(BCI2)), 10000)] <- 0
sp2 <- specaccum(BCI2, 'random')

Plotting

# Combine the specaccum objects into a list 
l <- list(sp1, sp2) 

# Calculate required y-axis limits
ylm <- range(sapply(l, '[[', 'richness') + 
           sapply(l, '[[', 'sd') * c(-2, 2))

# Apply a plotting function over the indices of the list
sapply(seq_along(l), function(i) {
  if (i==1) { # If it's the first list element, use plot()
    with(l[[i]], {
      plot(sites, richness, type='l', ylim=ylm, 
           xlab='Sites', ylab='random', las=1)
      segments(seq_len(max(sites)), y0=richness - 2*sd, 
               y1=richness + 2*sd)
    })    
  } else {
    with(l[[i]], { # for subsequent elements, use lines()
      lines(sites, richness, col=i)
      segments(seq_len(max(sites)), y0=richness - 2*sd, 
               y1=richness + 2*sd, col=i)
    })     
  }
})

legend('bottomright', c('Site 1', 'Site 2'), col=1:2, lty=1, 
       bty='n', inset=0.025)

enter image description here

Upvotes: 3

Related Questions