Shimodaira-Hasegawa Test
This function computes the Shimodaira–Hasegawa test for a set of trees.
SH.test(..., B = 10000, data = NULL, weight = NULL)
... |
either a series of objects of class |
B |
the number of bootstrap replicates. |
data |
an object of class |
weight |
if a matrix with site (log-)likelihoods is is supplied an optional vector containing the number of occurances of each site pattern. |
a numeric vector with the P-value associated with each tree given in
....
Klaus Schliep klaus.schliep@gmail.com
Shimodaira, H. and Hasegawa, M. (1999) Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Molecular Biology and Evolution, 16, 1114–1116.
data(Laurasiatherian) dm <- dist.logDet(Laurasiatherian) tree1 <- NJ(dm) tree2 <- unroot(upgma(dm)) fit1 <- pml(tree1, Laurasiatherian) fit2 <- pml(tree2, Laurasiatherian) fit1 <- optim.pml(fit1) # optimize edge weights fit2 <- optim.pml(fit2) # with pml objects as input SH.test(fit1, fit2, B=1000) # in real analysis use larger B, e.g. 10000 # with matrix as input X <- matrix(c(fit1$siteLik, fit2$siteLik), ncol=2) SH.test(X, weight=attr(Laurasiatherian, "weight"), B=1000) ## Not run: example(pmlPart) SH.test(sp, B=1000) ## End(Not run)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.