Phangorn bootstrap option

Question

[Speed.Lee]

I am using phangorn, maximum likelihood to estimate tree.

optim.pml process, I used rearrangement = "stochastic",

fit <-optim.pml(fit, multicore=TRUE, rearrangement = "stochastic")

so when bootstrap I should use optnni = TRUE or go straight forward? What is the difference, and the

bs <- bootstrap.pml(fit, bs=100, optnni = TRUE)

Sorry for the silly question but I cannot find an answer by search.

Answer 1

[Speed.Lee]

I just read a Phangorn vignettes, in appendix:

fit = optim.pml(fitNJ, rearrangement = "stochastic", optInv=TRUE,optGamma=TRUE)

fit

bs = bootstrap.pml(fit, bs=100, optNni=TRUE, multicore=TRUE)

This solves the problem.

But I still do not know the difference.