--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/biogeobears/b4bba74d-d390-4f9e-b977-49cae25b912d%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Hi Nick,I am also running BioGeoBEARS on a large dataset (900 terminals and 11 areas - maximum numer of occupied areas=9).On a linux cluster (120 cores - 64 GB memory) I manage to run part of DEC analysis. However, before it completes the analysis and loads results on .Rdata, I get this error message:Uppass starting for marginal ancestral states estimation!Error in if (sum(relprob_each_split_scenario) > 0) { :missing value where TRUE/FALSE neededCalls: bears_optim_run ... calc_uppass_probs_new2 -> calc_uppass_scenario_probs_new2In addition: Warning message:In (function (npt = min(n + 2L, 2L * n), rhobeg = NA, rhoend = NA, :unused control arguments ignoredExecution haltedCould you please tell me what does this exactly mean and how could I fix it?Thank you in advance for your help!Oscar
On Wednesday, July 8, 2015 at 1:21:40 AM UTC+2, chris blair wrote:Hi Nick,I am new to your program and am trying to analyze a relatively large data set consisting of 1277 species classified into 21 areas. The maximum number of areas occupied by a species is 16. I was hoping that you could provide some advice regarding parameter settings in BioGeoBEARS. I ran a preliminary unconstrained DEC analysis in RASP that finished in less than a day. However, I had to constrain the maximum number of areas to a very low value (e.g. 2). I am pretty sure that no analysis will finish if I specify a max areas value of 16. An obvious solution is to reduce the number of areas. However, we are a bit hesitant to do this as we would lose power. There is obviously a trade-off between power and resolution and being able to run a meaningful analysis to completion. Any thoughts are welcome.Chris
--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/biogeobears/52abfe40-6a02-4487-b79d-b6faf9ad6488%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/biogeobears/b98b21ff-69b1-418b-89fe-6bccef494136%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/biogeobears/6c571e10-7746-4744-b9e9-f04c68ac5a8c%40googlegroups.com.
Uppass starting for marginal ancestral states estimation!Error in if (sum(relprob_each_split_scenario) > 0) { :missing value where TRUE/FALSE neededCalls: bears_optim_run ... calc_uppass_probs_new2 -> calc_uppass_scenario_probs_new2In addition: Warning message:In (function (npt = min(n + 2L, 2L * n), rhobeg = NA, rhoend = NA, :unused control arguments ignoredExecution haltedMaybe is a silly question, but, do you think that this might be caused by very small branches on the tree?