I am doing a Biogeobears analysis across multiple bootstrap trees. I notice that there is a source code available for such purpose in BioGeoBEARS_on_multiple_trees_v1.R. I prepared all the input files, including one that contains the filenames of the tree files. I scan this file into the newick_fns argument as follows:
newick_fns = scan(“treelist.txt", what="character", sep=NULL)
And trying to run the run_bears_optim_on_multiple_trees function with following line:
res = run_bears_optim_on_multiple_trees(BioGeoBEARS_run_object, newick_fns, model_name="Elatinaceae_DEC", geog_fns, resfns, run_or_collect="run", start_treenum=1, end_treenum=length(newick_fns), runslow=TRUE, plot_params=FALSE)
As a result, it turns out I can only run the first tree. And when the analysis goes into the second tree, the following error occurred (this is a test run, so there are only three trees):
Could you please help me with that? In addition, is it correct to have each bootstrap tree separately in a file, and list the filename as a string in newick_fns? There’s also a function to summarize the result the plot the parameters onto a master tree. I think I need guide to use those functions as well: what the input argument should be like and how to call the function correctly in a R console.
ps. My input files and R scripts are attached here.
Thank you very much and best regards,
Liming
To view this discussion on the web visit https://groups.google.com/d/msgid/biogeobears/59d49dc9-933b-403d-a511-9ca807dfc5ad%40googlegroups.com.--
You received this message because you are subscribed to the Google Groups "BioGeoBEARS" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears...@googlegroups.com.
To post to this group, send email to bioge...@googlegroups.com.
Visit this group at http://groups.google.com/group/biogeobears.
Hi all,
I got the same error as Liming.
When I checked the objects res2$state_probs_at_nodes_across_all_trees and res2$state_probs_at_corners_across_all_trees I found out that they contain the probabilities of each node and each tree, not the average probabilities of all trees analyzed. These objects have n times (the number of trees analyzed) the number of rows of res$ML_marginal_prob_each_state_at_branch_top_AT_node (probabilities at nodes for the single tree analysis), I expected them to be the same size..
So maybe this is the problem? How can I obtain the averaged probabilities at nodes and corners?
Thanks,