xlinks allowed combinations
24 views
Skip to first unread message
Buz Wilson
Aug 12, 2024, 11:03:04 AM8/12/24
to TNT-Tree Analysis using New Technology, Markus Mendes-Grams
Dear TNT folk
I have sent this message as a bug report but perhaps someone else out there could advise me.
I am using TNT Version 1.6 - 64 bits - (January 2024) on my LINUX Xeon workstation with 40MB memory (2 processors, 16 cores, 32 threads) so I have plenty of computing resources available. Currently I am analyzing a subordinal level morphological dataset of extant and fossil Phreatoicidea and have been concerned about the quantity of inapplicables in the analysis; the family groups are extremely different but I want them all in the database because DELTA also outputs, in addition to the nexus file, natural language descriptions and keys and other taxonomic tools. Goloboff & Dr DeLaet's recent innovation of xlinks could be a boon for my phylogenetic analysis and a collaborative project on peracarids with colleagues in Rostock (Richter, Grams).
I finally have rough (=buggy) bash scripts to convert the dependencies in my database system (DELTA) into xlinks and am returning to the analysis of the Phreatoicidea. I ran into several problems, the first of which seems to be a bug.
Because of the many dependencies, some nested to 2 or more levels, both command line & Gui version of TNT output the error "Too many combinations in complex 11 (can't do more than 32767!) xmu; " Now, complex 11 is a simple unnested pair of characters, while complex 12 has many dependent characters, some nested. By removing one or the other from the analysis, I determined that the actual error should be referencing complex 12. I can send the datafile & most recent log file in case you need it. Please send the request to me directly to avoid unnecessary posts on the forum.
The second issue is the " (can't do more than 32767!) " problem. "xlinks = 32" can't be increased beyond this and may not be part of the problem.
Can the number of allowed combinations be changed by configuration or do I have to weed out the less consequential characters to bring it down to an acceptable value? In this instance, the program would be helpful if it would also estimate the number of combinations that caused the error, so the user could see how close one is to fixing the problem.
best wishes
Buz Wilson
I have sent this message as a bug report but perhaps someone else out there could advise me.
I am using TNT Version 1.6 - 64 bits - (January 2024) on my LINUX Xeon workstation with 40MB memory (2 processors, 16 cores, 32 threads) so I have plenty of computing resources available. Currently I am analyzing a subordinal level morphological dataset of extant and fossil Phreatoicidea and have been concerned about the quantity of inapplicables in the analysis; the family groups are extremely different but I want them all in the database because DELTA also outputs, in addition to the nexus file, natural language descriptions and keys and other taxonomic tools. Goloboff & Dr DeLaet's recent innovation of xlinks could be a boon for my phylogenetic analysis and a collaborative project on peracarids with colleagues in Rostock (Richter, Grams).
I finally have rough (=buggy) bash scripts to convert the dependencies in my database system (DELTA) into xlinks and am returning to the analysis of the Phreatoicidea. I ran into several problems, the first of which seems to be a bug.
Because of the many dependencies, some nested to 2 or more levels, both command line & Gui version of TNT output the error "Too many combinations in complex 11 (can't do more than 32767!) xmu; " Now, complex 11 is a simple unnested pair of characters, while complex 12 has many dependent characters, some nested. By removing one or the other from the analysis, I determined that the actual error should be referencing complex 12. I can send the datafile & most recent log file in case you need it. Please send the request to me directly to avoid unnecessary posts on the forum.
The second issue is the " (can't do more than 32767!) " problem. "xlinks = 32" can't be increased beyond this and may not be part of the problem.
Can the number of allowed combinations be changed by configuration or do I have to weed out the less consequential characters to bring it down to an acceptable value? In this instance, the program would be helpful if it would also estimate the number of combinations that caused the error, so the user could see how close one is to fixing the problem.
best wishes
Buz Wilson
-- George D.F. (Buz) Wilson, Ph.D. Saugatuck Natural History Laboratory, LLC publications: www researchgate net/profile/George-Wilson-12 email: gdfw at SNHLab com
Reply all
Reply to author
Forward
0 new messages