search

katylettuce / beast-mcmc

Automatically exported from code.google.com/p/beast-mcmc
0 stars 0 forks source link

invalid values during phylogeographic inference in continious space #742

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago 
For several weeks now I am trying to run a continuous phylogeography similar to 
the tutorial "Phylogeographic inference in continuous space" by  P. Lemey and 
N. Trovao (February 2013). But I am unable to obtain results using my data. My 
data consist of 48 taxa with 547 sites.  

I always end up with invalid values (see attached Screen Shot Tracer) at the 
end of the run. At the beginning of the run you can see the values for 
posterior en the prior change to "?" and will stay like this until the end of 
the run (see attached Screen Shot BEAST)
During some of my many failed runs I got the message:  "Underflow calculating 
like likelihood... attempting a rescaling" but this is not always the case. 

I have partitioned my data to a smaller number of taxa (16) to see if the 
number of taxa were the problem. Out of 10 independent runs I have been able to 
obtain 1 result with valid values but with a very low ESS. So results of 
different runs are also inconsistent...

After trying many different models and combinations I do not see where the 
problem is. And the worst thing is that I have no problems running the data 
provided within the tutorial.

Has anyone seen this problem before? Or can someone please advise me on how to 
tackle these invalid values?

I am using BEAST v.1.8.0 on either a Mac Os X or Ubuntu. In both operating 
systems I have the same problem.

Original issue reported on code.google.com by willemse...@gmail.com on 4 Aug 2014 at 3:57

Attachments:

GoogleCodeExporter commented 9 years ago 
This probably results from having tips with identical locations in the 
continuous diffusion model. This should probably be checked for by BEAST so I 
will add this as an issue and flag this as invalid. Problems like this are best 
posted to the mailing list.

Original comment by ramb...@gmail.com on 9 Aug 2014 at 2:52

GoogleCodeExporter commented 9 years ago 
Thank you for your reply. I have sent an email 2 months ago but I have not 
received any answer. Can you please provide me with the email addresses where I 
can post this problem. I would like to have it solved, if possible. Thank you 
very much in advance

Original comment by willemse...@gmail.com on 11 Aug 2014 at 9:25