Description Usage Arguments Details Value Note Author(s) References See Also Examples
This function constructS eigenvectors of a site-by-link matrix. Weights can be applied to the links.
1 | aem(build.binary, binary.mat, weight, rm.link0 = FALSE, print.binary.mat = FALSE)
|
build.binary |
Object created by function |
binary.mat |
Site (n rows) by link (k columns) matrix. 1 represents the presence of a link influencing a site, directly or indirectly, otherwise 0. |
weight |
Vector of weights of length k, to be applied to the links. |
rm.link0 |
Logical ( |
print.binary.mat |
Logical ( |
If only an object of class build.binary
is given to this function, The argument binary.mat
is not considered. binary.mat
is only considered when the argument build.binary
is missing.
If weights are applied to the links, the length of vector weight
has to take into account wether the links connecting real sites to the origin (the fictitious site 0) have been kept or removed.
value |
A vector of singular values associated with the AEM. |
vectors |
A matrix of eigenvector. Each column is an AEM eigenfunction (or variable). |
mod.binary.mat |
A site-by-link matrix modified through the function. |
It sometimes happens that AEM eigenfunctions have equal singular values.In that case, different sets of AEM eigenfunctions may be produced on different plateforms. Eigenvectors associated to an eigenvalue that is smaller than 10^{-12} are considered negligeable. They have been removed from the created AEM eigenfunctions.
F. Guillaume Blanchet
Blanchet F.G., P. Legendre, and Borcard D. 2008, Modelling directional spatial processes in ecological data. Ecological Modelling, 215, 325-336.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 | ### Construction of object of class nb (spdep)
nb<-cell2nb(5,5,"queen")
### Create fictitious geographical coordinates
xy <- cbind(1:25,expand.grid(1:5,1:5))
### Build binary site-by-link matrix
bin.mat <- build.binary(nb,xy)
### Construct AEM eigenfunctions from an object of class build.binary
res <- aem(build.binary=bin.mat,rm.link0=FALSE)
res$values
### Illustrate 4 AEM eigenfunctions using bubble plots
opal <- palette()
palette(c("black","white"))
par(mfrow=c(2,2))
symbols(x=xy[,2:3],circles=abs(res$vectors[,1]),inches=FALSE,asp=1,fg=ifelse(sign(-res$vectors[,1])+1>0,1,0),bg=ifelse(sign(res$vectors[,1])+1>0,1,0),xlab="x",ylab="y")
title("AEM 1")
symbols(x=xy[,2:3],circles=abs(res$vectors[,2]),inches=FALSE,asp=1,fg=ifelse(sign(-res$vectors[,2])+1>0,1,0),bg=ifelse(sign(res$vectors[,2])+1>0,1,0),xlab="x",ylab="y")
title("AEM 2")
symbols(x=xy[,2:3],circles=abs(res$vectors[,3]),inches=FALSE,asp=1,fg=ifelse(sign(-res$vectors[,3])+1>0,1,0),bg=ifelse(sign(res$vectors[,3])+1>0,1,0),xlab="x",ylab="y")
title("AEM 3")
symbols(x=xy[,2:3],circles=abs(res$vectors[,4]),inches=FALSE,asp=1,fg=ifelse(sign(-res$vectors[,4])+1>0,1,0),bg=ifelse(sign(res$vectors[,4])+1>0,1,0),xlab="x",ylab="y")
title("AEM 4")
### Construct AEM eigenfunctions using only a site-by-link matrix
res2 <- aem(binary.mat=bin.mat[[1]])
res2$values
### Illustrate 4 AEM eigenfunctions using bubble plots
par(mfrow=c(2,2))
symbols(x=xy[,2:3],circles=abs(res2$vectors[,1]),inches=FALSE,asp=1,fg=ifelse(sign(-res2$vectors[,1])+1>0,1,0),bg=ifelse(sign(res2$vectors[,1])+1>0,1,0),xlab="x",ylab="y")
title("AEM 1")
symbols(x=xy[,2:3],circles=abs(res2$vectors[,2]),inches=FALSE,asp=1,fg=ifelse(sign(-res2$vectors[,2])+1>0,1,0),bg=ifelse(sign(res2$vectors[,2])+1>0,1,0),xlab="x",ylab="y")
title("AEM 2")
symbols(x=xy[,2:3],circles=abs(res2$vectors[,3]),inches=FALSE,asp=1,fg=ifelse(sign(-res2$vectors[,3])+1>0,1,0),bg=ifelse(sign(res2$vectors[,3])+1>0,1,0),xlab="x",ylab="y")
title("AEM 3")
symbols(x=xy[,2:3],circles=abs(res2$vectors[,4]),inches=FALSE,asp=1,fg=ifelse(sign(-res2$vectors[,4])+1>0,1,0),bg=ifelse(sign(res2$vectors[,4])+1>0,1,0),xlab="x",ylab="y")
title("AEM 4")
palette(opal)
### Construct AEM eigenfunctions with a function of the distance as weights to put on the links
### Construction of object of class nb (spdep)
nb<-cell2nb(5,5,"queen")
### Create fictitious geographical coordinates
xy <- cbind(1:25,expand.grid(1:5,1:5))
### Build binary site-by-link matrix
bin.mat <- build.binary(nb,xy)
### Construct a matrix of distances
long.lien.mat<-as.matrix(dist(xy))
### Extract the edges, remove the ones directly linked to site 0
lien.b<-bin.mat$edges[-1:-5,]
### Construct a vector giving the length of each edge
long.lien<-vector(length=nrow(lien.b))
for(i in 1:nrow(lien.b)){
long.lien[i]<-long.lien.mat[lien.b[i,1],lien.b[i,2]]
}
### Construct a vector of weights based on distance
weight.vec<-1-(long.lien/max(long.lien))^2
### Construct AEM eigenfunctions from an object of class build.binary
res <- aem(build.binary=bin.mat,weight=weight.vec,rm.link0=TRUE)
res
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.