You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I use a freshly compiled version of PhyML cloned from the GitHub repository (date: 19/07/15) and enabled the BEAGLE support for speeding up the likelihood calculations.
When running the BEAGLE-enabled PhyML with setting the "Tree topology search operations" to "Best of NNI and SPR" and 5 random starting trees, all of my data sets crash at some point with the following error code:
. Best_move < 0 !
. List size = 2
. 0x12c49c30 0x242d9b0
. 0x12c3c2b0 0x23fdc50
== Err. in file spr.c at line 4166 (function 'Evaluate_List_Of_Regraft_Pos_Triple')
The list size is varying based on the data set and the time when crashing. I cannot reproduce this behaviour when using the example data set delivered with PhyML. But the sample size in my data files is much larger and the alignment length much longer (>300 samples, alignment length > 2,000bp). BEAGLE itself should be the problem because I am able to run BEAST using the same library without any problems.
Is this behaviour due to my data sets (nucleotide alignments with gaps) or due to a bug in PhyML's spr.c?
Thanks in advance for your help,
Alex
The text was updated successfully, but these errors were encountered:
Hello,
I use a freshly compiled version of PhyML cloned from the GitHub repository (date: 19/07/15) and enabled the BEAGLE support for speeding up the likelihood calculations.
When running the BEAGLE-enabled PhyML with setting the "Tree topology search operations" to "Best of NNI and SPR" and 5 random starting trees, all of my data sets crash at some point with the following error code:
. Best_move < 0 !
. List size = 2
. 0x12c49c30 0x242d9b0
. 0x12c3c2b0 0x23fdc50
== Err. in file spr.c at line 4166 (function 'Evaluate_List_Of_Regraft_Pos_Triple')
The list size is varying based on the data set and the time when crashing. I cannot reproduce this behaviour when using the example data set delivered with PhyML. But the sample size in my data files is much larger and the alignment length much longer (>300 samples, alignment length > 2,000bp). BEAGLE itself should be the problem because I am able to run BEAST using the same library without any problems.
Is this behaviour due to my data sets (nucleotide alignments with gaps) or due to a bug in PhyML's spr.c?
Thanks in advance for your help,
Alex
The text was updated successfully, but these errors were encountered: