Excluding positions and charpartitions

[Britta]

Dear Derrick,

I have a big alignment and I want to calculate several trees from several subsets of this. Therefore I want to exclude data.

I do it like this for two examples
begin sets;
charset Gene1_pos1 = 1-418\3;
charset Gene1_pos2 = 2-418\3;
charset Gene1_pos3 = 3-418\3;
charpartition Position = chunk1:Gene1_pos1 Gene1_pos2, chunk2:Gene1_pos3;
end;
begin assumptions;
  exset* badsites = 419-17545;
end;
AND like this

begin sets;
charset Gene33_pos1 = 12963-13053\3;
charset Gene33_pos2 = 12964-13053\3;
charset Gene33_pos3 = 12965-13053\3;
charset intron33 = 13054-13487;
charpartition Position = chunk1:Gene33_pos1 Gene33_pos2 Gene33_pos3 , chunk2:intron33;
end;
begin assumptions;
  exset* badsites = 1-12962 13488-17545 ;
end;


I get this warning:
Warning: 
 Position is a not a valid CharPartition. At least one Character (1) is not included
at line 13247, column (approximately) 88 (file position 803730)
storing read block: SETS
 successful
Reading ASSUMPTIONS block...storing read block: ASSUMPTIONS
 successful

Why does it give that warning? Could you explain if I did something wrong? Should I be concerned?

Thanks a lot,

Britta

[Derrick Zwickl]

Hi Britta,

I see why that message would be confusing.  The basic reason is that by definition a charpartition should include all characters (that haven't been deleted), but the check of whether this is the case is done BEFORE the exset is processed.  So, when it warns you, it doesn't yet know that you've excluded all of the characters that aren't in the charpartition.  I don't think that I've tested that sort of scenario extensively, but it shouldn't be a problem.  To be sure, carefully check the output regarding the number of characters in each partition subset and verify that it matches what you specified.

If you run into problems or other confusion, let me know.

Best,
Derrick