plot.fpcomSims: Plot fpcomSims objects

Description Usage Arguments Details Value


This plot method can produce four different graphical summaries of the output of the fpcomSims function.


## S3 method for class 'fpcomSims'
plot(x, y, cex.add = 0.8, plottype = c("distance",
  "traitgram", "gradient", "ordination"), ...)



An fpcomSims object.


Not used at the moment.


The cex graphics parameter for additional graphical. elements not produced by the plot command.


The type of plot to produce – see details.


Additional parameters to be passed to plot.


plottype equals "distance"

A scatterplot of the phylogenetic versus functional distances between the species pairs are produced. Species numbers are used to label the points. The cophenetic function is used to compute the phylogenetic distances and the dist function is used to compute the functional Euclidean distances. Both distances are standardised such that the maximum distance is one.

plottype equals "traitgram"

A traitgram is produced, which combines both the phylogeny and the observed trait.

plottype equals "gradient"

A scatterplot of the probabilities of occurrence versus the observed gradient is produced. The species are identified by their numbers. Note that these probabilities of occurrence depend only on the two gradients and the two traits, and therefore are not subject to variation among sites with identical gradient values. Such variation is expressed in the comm element of fpcomSims objects, which gives not probability of occurrence but occurrence itself. Note that if the p argument to fpcomSims is one, then only the observed trait and gradient determine probability of occurrence and the plot consists of perfect sigmoid curves. But if p is zero then only the unknown gradient and trait determine probability of occurrence and the plot is just noise. In this latter case we would expect phylogenetic distance to provide more important information about communities.

plottype equals "ordination"

A special type of ordination of the species is produced. This ordination is based on the probabilities of occurrence and not on occurrence itself, and so it is able to fully explain all variation on two axes (if p is not either zero of one). Therefore, such an ordination is not possible in practice with real data but it is useful in this case for fully representing distances between species in species distribution space. Technically it is a singular value decomposition of the logit transformed probabilities of occurrence. Note well the percentages of variation explained by the two axes, and that this ordination is gauranteed to have 100 percent of the variation explained by the first two axes.


No return value, called for its side-effect of producing a plot.

funphylocom documentation built on May 2, 2019, 5 p.m.