ptau6: ptau6

In deformetrics/deformetrics: R functions for shape analysis

Description

This function perform the Euclidean Parallel transport for allometric investigations using the Linear Shift startegy described in Piras et al (2016). The Linear Shift is performed on MANCOVA predictions and original data are projected on PCA space computed on them.

Usage

ptau6(array, factor, CSinit = T, sepure = F, polyn = 1, CR = NULL,
  locs = NULL, perm = 999)

Arguments

array	numeric: an array kxmxn of landmark coordinates
factor	character: variable factor that affiliates shapes to group levels
CSinit	logical: if TRUE shapes are scaled at unit size (default=TRUE)
sepure	logical: if TRUE separate per-group multivariate regression between shape and size are performed on shapes aligned after separate GPAs (default=FALSE)
polyn	numeric: default=1 the degree of regression
perm	numeric: number of permutations for group non parametric regression

Author(s)

Paolo Piras

References

Piras P., Teresi L., Traversetti L., Varano V, Gabriele S., Kotsakis T., Raia P., Puddu P.E., Scalici M. (2016). The conceptual framework of ontogenetic trajectories: Parallel Transport allows the recognition and visualization of pure deformation patterns. Evolution and Development 18: 182-200. doi: 10.1111/ede.12186

Examples

## Not run:  
library(Morpho)
library(gdata)
library(rgl)
data(group)
data(my2d)
mysel<-group%in%c("Macroscelides_proboscideus","Petrodromus_tetradactylus","Elephantulus_rozeti","Elephantulus_edwardii")
linksdors<-list(c(1,2),c(37,7),c(12,4),c(27,28),c(25,21),c(38,40),c(9,10),c(2,3),c(3,4),c(1,7),c(1,6),c(3,5),c(6,40),c(5,9),c(40,8),c(8,9),c(1,7),c(7,6),c(3,4),c(4,5),c(39,38),c(38,35),c(35,37),c(37,39),c(35,34),c(34,33),c(33,32),c(32,31),c(31,30),c(30,29),c(29,37),c(37,36),c(36,29),c(28,31),c(28,30),c(13,10),c(10,11),c(11,12),c(12,13),c(13,14),c(14,16),c(16,17),c(17,20),c(20,19),c(19,18),c(18,12),c(18,15),c(15,12),c(21,19),c(21,20),c(24,25),c(25,26),c(26,27),c(27,24),c(26,24),c(24,23),c(23,22),c(22,8),c(8,2))       
dors4<-my2d[,,mysel]
factordors4<-drop.levels(group[mysel],reorder=T)
factordors4<-factor(factordors4,levels=unique(factordors4))
adors4<-procSym(dors4,scale=F,pcAlign=F,reflect=F)
mypredictbook<-read.inn(predict(lm(array2mat(adors4$orpdata,105,80)~poly(adors4$size,1,raw=T)*factordors4)),40,2)
procbook<-procSym(mypredictbook,pcAlign=F)
# linear shift in the shape space
plot(procbook$PCscores[,1:2],pch=as.numeric(factordors4),asp=1)
objptau<-ptau6(dors4,factordors4,CSinit=T)
#  morphological apprciation is dramatically different!
plotptau5(objptau,linksdors,pch=as.numeric(objptau$factorord),col=rep(1,length(objptau$factorord)),round(max(centroid.size(objptau$arrayord)),digits=0)/2,shiftnegy=2,mag=2,subplotdim=1)
## End(Not run)

## End(Not run)

deformetrics/deformetrics documentation built on May 15, 2019, 3:20 a.m.