Mapping morphological convergence on 3D surfaces
Given vectors of RW (or PC) scores, the function selects the RW(PC) axes which best account for convergence and maps convergent areas on the corresponding 3D surfaces.
conv.map(dataset,pcs,mshape,conv=NULL, exclude=NULL,out.rem=TRUE, show.consensus=FALSE, plot=TRUE,col="blue",names = TRUE)
dataset |
data frame (or matrix) with the RW (or PC) scores of the group or species to be compared. |
pcs |
RW (or PC) vectors (eigenvectors of the covariance matrix) of all the samples. |
mshape |
the Consensus configuration. |
conv |
a named character vector indicating convergent species as
(indicated as "conv" in |
exclude |
integer: the index number of the RW (or PC) to be excluded from the comparison. |
out.rem |
logical: if |
show.consensus |
logical: if |
plot |
logical: if |
col |
character: the colour for the plot. |
names |
logical: if |
conv.map automatically builds a 3D mesh on the mean shape
calculated from the Relative Warp Analysis (RWA) or Principal Component
Analysis (PCA) (Schlager 2017) by applying the function
vcgBallPivoting (Rvcg). conv.map further gives
the opportunity to exclude some RW (or PC) axes from the analysis because,
for example, in most cases the first axes are mainly related to high-order
morphological differences driven by phylogeny and size variations.
conv.map finds and plots the strength of convergence on 3D surfaces.
An output of conv.map (if the dataset contains a number equal or
lower then 5 items) is an interactive plot mapping the convergence on the
3D models. In the upper triangle of the 3D multiple layouts the rows
representing the reference models and the columns the target models. On the
contrary, on the lower triangle the rows correspond to the target models
and the columns the reference models. In the calculation of the differences
of areas we supply the possibility to find and remove outliers from the
vectors of areas calculated on the reference and target surfaces. We
suggest considering this possibility if the mesh may contain degenerate
facets.
The function returns a list including:
$angle.compare data frame including the real angles between the given shape vectors, the angles conv computed between vectors of the selected RWs (or PCs), the angles between vectors of the non-selected RWs (or PCs), the difference conv, and its p values.
$selected.pcs RWs (or PCs) axes selected for convergence.
$average.dist symmetric matrix of pairwise distances between 3D surfaces.
$suface1 list of coloured surfaces, if two meshes are given, it represents convergence between mesh A and B charted on mesh A.
$suface2 list of coloured surfaces, if two meshes are given, it represents convergence between mesh A and B charted on mesh B.
$scale the value used to set the colour gradient, computed as the maximum of all differences between each surface and the mean shape.
Marina Melchionna, Antonio Profico, Silvia Castiglione, Carmela Serio, Gabriele Sansalone, Pasquale Raia
Schlager, S. (2017). Morpho and Rvcg–Shape Analysis in R: R-Packages for geometric morphometrics, shape analysis and surface manipulations. In: Statistical shape and deformation analysis. Academic Press.
## Not run:
data(DataSimians)
DataSimians$pca->pca
## Case 1. Convergent species only
dato<-pca$PCscores[c(1,4),]
CM<-conv.map(dataset = dato,
pcs = pca$PCs,
mshape = pca$mshape,
show.consensus = TRUE)
## Case 2. Convergent and non-convergent species
dato<-pca$PCscores[c(1,4,7),]
conv<-c("conv","conv","noconv")
names(conv)<-rownames(dato)
CM<-conv.map(dataset = dato,
pcs = pca$PCs,
mshape = pca$mshape,
conv = conv,
show.consensus = TRUE,
col = "orange")
## End(Not run)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.