I've been trying to delimitate a set of 18 species, 6 already described and the rest cryptic with BPP. I am running A11 analysis on 162 specimens with 4 genes for 50000 generations. My analysis run fine, but when the software needs to summarize mcmc it collapses every time. Original BPP just stop, so I additonaly tried to run it with bppX. BppX gives me following Error:
Error: oom models || countmodel.
read tree 40000
does it mean that too many trees or models are included. Just to remind, 18 species designations gives 300+ possible delimitation options. Thank you for any kind of help.
Sincerely,
Teo
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
oom means out of memory.
Possibly the information content in the data is low but anyway the posterior may be diffuse and the chain visited many distinct models and when the program tries to summarise the models they are all read into the memory somehow. The memory requirement is not proportional to nsample but is proportional to the number of distinct models in the sample.
It may be possible to redesign the code to reduce the memory intake, since the code was written in a hurry, but that won't happen anytime soon.
You can try to separate the sample file into two halves and summarise each half and see whether that works. Then take an average over the two halves.
Note that print=-1 summarises the sample without running the MCMC.
One other possibility may be to prepare two smaller datasets and analyze them separately, to reduce the number of possible models.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
View and moderate all "BP&P Q&A" comments posted by this user
Mark all as spam, and block user from posting to "Discussion"
Dear Users,
I've been trying to delimitate a set of 18 species, 6 already described and the rest cryptic with BPP. I am running A11 analysis on 162 specimens with 4 genes for 50000 generations. My analysis run fine, but when the software needs to summarize mcmc it collapses every time. Original BPP just stop, so I additonaly tried to run it with bppX. BppX gives me following Error:
Error: oom models || countmodel.
read tree 40000
does it mean that too many trees or models are included. Just to remind, 18 species designations gives 300+ possible delimitation options. Thank you for any kind of help.
Sincerely,
Teo
Yang says:
oom means out of memory.
Possibly the information content in the data is low but anyway the posterior may be diffuse and the chain visited many distinct models and when the program tries to summarise the models they are all read into the memory somehow. The memory requirement is not proportional to nsample but is proportional to the number of distinct models in the sample.
It may be possible to redesign the code to reduce the memory intake, since the code was written in a hurry, but that won't happen anytime soon.
You can try to separate the sample file into two halves and summarise each half and see whether that works. Then take an average over the two halves.
Note that print=-1 summarises the sample without running the MCMC.
One other possibility may be to prepare two smaller datasets and analyze them separately, to reduce the number of possible models.
View and moderate all "BP&P Q&A" comments posted by this user
Mark all as spam, and block user from posting to "Discussion"
Thank you for quick anwser
teo