be clearer about when abundance weighting is used in gather output

[ctb]

I was looking at sourmash gather output and trying to figure out how this metagenome was 99.2% classified:

== This is sourmash version 4.2.4.dev0+g73aeb155.d20220116. ==
== Please cite Brown and Irber (2016), doi:10.21105/joss.00027. ==

selecting default query k=31.
loaded query: SRR12324253... (k=31, DNA)
loaded 1 databases.                                                            

Starting prefetch sweep across databases.
Found 18 signatures via prefetch; now doing gather.

overlap     p_query p_match avg_abund
---------   ------- ------- ---------
15.8 Mbp       0.4%   62.7%       2.6    SSXJ01000001.1 Cryptococcus neoforman...
11.3 Mbp       0.7%   97.6%       7.4    WMJW01000001.1 Saccharomyces cerevisi...
7.4 Mbp       30.0%   67.7%     461.6    JSPL01000060.1 Escherichia coli strai...
4.7 Mbp       24.7%   98.4%     607.9    VFAF01000002.1 Salmonella enterica st...
4.0 Mbp        2.8%   99.8%      80.1    CP039755.1 Bacillus subtilis strain N...
5.2 Mbp       13.3%   43.3%     474.1    FREP01002036.1 Escherichia coli isola...
2.8 Mbp        1.7%  100.0%      68.6    CP039751.1 Listeria monocytogenes str...
2.7 Mbp        4.6%   99.8%     192.8    VFAE01000004.1 Staphylococcus aureus ...
10.3 Mbp       0.0%   10.0%       2.2    AE017341.1 Cryptococcus neoformans va...
1.8 Mbp        6.2%   99.8%     395.3    CP039750.1 Limosilactobacillus fermen...
6.8 Mbp        6.2%   26.3%     394.4    CP041013.1 Pseudomonas aeruginosa str...
8.4 Mbp        0.0%    9.1%       1.9    CP003820.1 Cryptococcus neoformans va...
2.8 Mbp        3.5%   36.0%     391.4    CP039752.1 Enterococcus faecalis stra...
4.7 Mbp        5.0%   20.0%     604.4    VFAG01000002.1 Escherichia coli strai...
10.6 Mbp       0.1%    2.4%      16.5    LOQK01000001.1 Saccharomyces pastoria...
0.5 Mbp        0.0%    9.6%       1.3    VEMH01000100.1 Bacillus paranthracis ...
10.3 Mbp       0.0%    2.5%      10.7    CM010209.1 Saccharomyces cerevisiae s...
10.4 Mbp       0.0%    1.3%       3.8    CM006175.1 Saccharomyces cerevisiae s...
found less than 50.0 kbp in common. => exiting

found 18 matches total;
the recovered matches hit 99.2% of the query

when I realized that it was only 99.2% classified when abundance weighting was used - with --ignore-abundance, the percentage drops to 42.9%.

selecting default query k=31.
loaded query: SRR12324253... (k=31, DNA)
loaded 1 databases.                                                            

Starting prefetch sweep across databases.
Found 18 signatures via prefetch; now doing gather.

overlap     p_query p_match
---------   ------- -------
15.8 Mbp      10.9%   62.7%    SSXJ01000001.1 Cryptococcus neoforman...
11.3 Mbp       7.8%   97.6%    WMJW01000001.1 Saccharomyces cerevisi...
7.4 Mbp        5.1%   67.7%    JSPL01000060.1 Escherichia coli strai...
4.7 Mbp        3.2%   98.4%    VFAF01000002.1 Salmonella enterica st...
4.0 Mbp        2.7%   99.8%    CP039755.1 Bacillus subtilis strain N...
5.2 Mbp        2.2%   43.3%    FREP01002036.1 Escherichia coli isola...
2.8 Mbp        1.9%  100.0%    CP039751.1 Listeria monocytogenes str...
2.7 Mbp        1.9%   99.8%    VFAE01000004.1 Staphylococcus aureus ...
10.3 Mbp       1.3%   10.0%    AE017341.1 Cryptococcus neoformans va...
1.8 Mbp        1.2%   99.8%    CP039750.1 Limosilactobacillus fermen...
6.8 Mbp        1.2%   26.3%    CP041013.1 Pseudomonas aeruginosa str...
8.4 Mbp        1.1%    9.1%    CP003820.1 Cryptococcus neoformans va...
2.8 Mbp        0.7%   36.0%    CP039752.1 Enterococcus faecalis stra...
4.7 Mbp        0.6%   20.0%    VFAG01000002.1 Escherichia coli strai...
10.6 Mbp       0.5%    2.4%    LOQK01000001.1 Saccharomyces pastoria...
0.5 Mbp        0.4%    9.6%    VEMH01000100.1 Bacillus paranthracis ...
10.3 Mbp       0.2%    2.5%    CM010209.1 Saccharomyces cerevisiae s...
10.4 Mbp       0.1%    1.3%    CM006175.1 Saccharomyces cerevisiae s...
found less than 50.0 kbp in common. => exiting

found 18 matches total;
the recovered matches hit 42.9% of the query

between that and some of @drtamermansour's experiences here I think we need to be clearer about these numbers...

on the plus side, it's pretty clear in this particular mock metagenome's case that most of the low-abundance k-mers are errors! (this is zymo mock, so is a low complexity mock community)

[ctb]

fixed in #1819.