50% probability at each node
Opened this issue · 2 comments
Hi Steven,
Thank you for this great program. I'm using it for the first time now and I'm facing the following issue. I have a phylogeny with about 180 spp. I have just one trait that can have two states (files attached). The trait is not ordered (either social or non-social). When I run the function, I obtain a tree that shows a 50% probability at each node. I might be missing something but not sure what.
Could you help me? Thank you in advance!
The NWK tree comes in TXT format so I could upload it :)
file_trait.txt
plot.settings.txt
tree_PB_good.nwk.txt
Pol Sorigué
Hi Pol,
Thanks for pointing this out. I left your tree and trait data files as you sent them, but I adjusted your plot settings file to make the output pdf a little more pleasing. I got a result similar to what you described when I re-ran it using 1,000 samples. So, I increased to 10,000 samples, re-ran again, and then looked at the auto-discard to see what's happening. I have included that for you to see (gens_v_likelihood.pdf).
It seems that the complexity of your tree and tip data should require at least 5,000 samples. Only samples to the right of the red vertical line AND ALSO above the red horizontal line are summarized in your result. I included the commands that I pasted into R so you can try to recreate. From what I can tell, requesting 5k or 10k samples will give you a very similar outcome. So if you are in a rush, 5k samples is probably fine.
Please let me know if you encounter any other issues.
All my best,
Steven