Known Issues and Bugs

Structure v.2.3.1, April 2009

• Polyploids: If you are analyzing polyploid data then you might encounter problems with Structure stalling and not returning results. This is a limitation of the current implementation of our sampling function. This problem occurs mainly for loci with very large numbers of alleles, and especially in data with high ploidy. The next release will partially fix the problem. In the meantime you might consider removing the most variable loci, or merging rare alleles to get the program to run.

• Windows with Frontend: Occasionally you may encounter a memory problem when trying to start the Java virtual machine and find the error message "Cannot create Java machine". This is because Structure is requesting more memory than what is available. As a result the user can run structure_start.bat which will fix the memory usage at 512Mb. If you choose to increase the memory usage you can edit this file and change the memory argument to an appropriate value based on your computer's resources.
• Windows with Frontend & Console: In rare cases, earlier versions of Structure produced an overflow/underflow bug which resulted in output that looked like *"Estimated Ln Prob of Data = -1.$"*, "Mean value of ln likelihood = -1.$" and "Variance of ln likelihood = -1.$". We believe that this problem has been fixed in Version 2.3.1