1. Analysis of alpha diversity
1.1. Benchmarking phyloregion for analysis of
phylogenetic diversity
For analysis of alpha diversity commonly used in conservation such as
phylogenetic diversity - the sum of all phylogenetic branch lengths
within an area (Faith 1992) -
phyloregion is 31 to 284 times faster and 67 to 192 times
memory efficient, compared to other packages!
tree <- africa$phylo
tree <- keep.tip(tree, colnames(X_sparse))
pd_picante <- function(x, tree){
res <- picante::pd(x, tree)[,1]
names(res) <- row.names(x)
res
}
pd_pez <- function(x, tree){
dat <- pez::comparative.comm(tree, x)
res <- pez::.pd(dat)[,1]
names(res) <- row.names(x)
res
}
pd_hilldiv <- function(x, tree) hilldiv::index_div(x, tree, index="faith")
pd_phyloregion <- function(x, tree) phyloregion::PD(x, tree)
res1 <- bench::mark(picante=pd_picante(X_dense, tree),
hilldiv=pd_hilldiv(Xt_dense,tree=tree),
pez=pd_pez(X_dense, tree),
phyloregion=pd_phyloregion(X_sparse, tree))
summary(res1)## # A tibble: 4 × 6
## expression min median `itr/sec` mem_alloc `gc/sec`
## <bch:expr> <bch:tm> <bch:tm> <dbl> <bch:byt> <dbl>
## 1 picante 97.85ms 112.45ms 8.22 59.6MB 9.87
## 2 hilldiv 762.42ms 762.42ms 1.31 170.1MB 3.93
## 3 pez 108.59ms 122.83ms 8.16 60.4MB 8.16
## 4 phyloregion 2.82ms 3.23ms 268. 909.8KB 5.99autoplot(res1)
plot of chunk phylo_diversity
1.2. Benchmarking phyloregion for analysis of
phylogenetic endemism
Another benchmark for phyloregion is in the analysis of
phylogenetic endemism, the degree to which phylogenetic diversity is
restricted to any given area (Rosauer et al.
2009). Here, we found that phyloregion is 160 times
faster and 489 times efficient in memory allocation.
tree <- africa$phylo
tree <- keep.tip(tree, colnames(X_sparse))
pe_pez <- function(x, tree){
dat <- pez::comparative.comm(tree, x)
res <- pez::pez.endemism(dat)[,1]
names(res) <- row.names(x)
res
}
pe_phyloregion <- function(x, tree) phyloregion::phylo_endemism(x, tree)
res2 <- bench::mark(pez=pe_pez(X_dense, tree),
phyloregion=pe_phyloregion(X_sparse, tree))
summary(res2)## # A tibble: 2 × 6
## expression min median `itr/sec` mem_alloc `gc/sec`
## <bch:expr> <bch:tm> <bch:tm> <dbl> <bch:byt> <dbl>
## 1 pez 630.38ms 630.38ms 1.59 493.88MB 9.52
## 2 phyloregion 2.94ms 3.19ms 280. 1.06MB 3.98autoplot(res2)
plot of chunk phylo_endemism
