summary stats vs actual alignments

[wbsimey]

Hello,
I am getting different results from the summary stats code vs phyluce_align_get_only_loci_with_min_taxa code. Your tutorial shows the same numbers for both at 75% matrix, 913 alignments.
Is this expected? I noticed that you use the ".floor" method in the phyluce_align_get_only_loci_with_min_taxa code. Perhaps this is the difference?

The following table holds my numbers for ABySS. The velvet assemblies are even further off (I can't get Trinity to work on Ubuntu 16 with Java 1.8 even after installing the latest Trinity and editing python.conf).

Matrix %	summary stats	phyluce_get_loci
65%	914 alignments	847 alignments
75%	720 alignments	551 alignments
85%	398 alignments	194 alignments

[brantfaircloth]

The numbers in summary stats are estimates (and the floor is what makes the difference). I need to fix this at some point...