search

lukejharmon / ilhabela

R course in Ilhabela, Brazil, June 2015
7 stars 7 forks source link

small typo fix: OU vs Pagel #1

Open cecileane opened 4 years ago

cecileane commented 4 years ago

Hi Luke,

I've seen people use corPagel instead of corMartins after following your instructions here (at the bottom). It looks like the confusion I've seen is from a simple typo in the text above your code, line 95 of PGLS.Rmd. Just replace "follows an OU model rather than Brownian motion" by "follows Pagel's lambda model rather than Brownian motion" (say).

And then add a comment within the code block below, line 108 saying that now we can use an error structure assuming the OU model.

Small things. For folks who use your nice tutorial.

r03ert0 commented 2 years ago

Thank you @cecileane for the warning! Is it the case then that Martins & Hansen model is equivalent to the Ornstein-Uhlenbeck model?

cecileane commented 2 years ago

Yes indeed. The function corMartins implements the correlation structure for the OU process, when the ancestral state at the root is modeled as a random variable with the stationary distribution of the OU process. It's the same as the "OUrandomRoot" correlation model of the phylolm function.