ignore specific characters in all distance calculations

[benjaminotter]

Description of proposed changes

Adds support for ignoring specific characters during distance calculation. The ignored characters are defined in the distance map as a list of characters:

{
    "default": 1,
    "ignored_characters": ["N", "-"],
    "map": {}
}

The specification of those characters in the distance map is optional, making it compatible with distance maps that don't specify any characters to ignore.

Related issue(s)

Fixes #693

Testing

Added doctests as suggested in #693:

Ignore specific characters defined in the distance map.

>>> node_a_sequences = {"gene": "ACTGG"}
>>> node_b_sequences = {"gene": "A--GN"}
>>> distance_map = {"default": 1, "ignored_characters":["-"], "map": {}}
>>> get_distance_between_nodes(node_a_sequences, node_b_sequences, distance_map)
1
>>> distance_map = {"default": 1, "ignored_characters":["-", "N"], "map": {}}
>>> get_distance_between_nodes(node_a_sequences, node_b_sequences, distance_map)
0

[codecov]

Codecov Report

Merging #707 (328c1ff) into master (5c12fd8) will increase coverage by 0.01%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #707      +/-   ##
==========================================
+ Coverage   30.89%   30.90%   +0.01%     
==========================================
  Files          41       41              
  Lines        5648     5649       +1     
  Branches     1365     1365              
==========================================
+ Hits         1745     1746       +1     
  Misses       3828     3828              
  Partials       75       75              

Impacted Files	Coverage Δ
augur/distance.py	37.79% <100.00%> (+0.49%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5c12fd8...328c1ff. Read the comment docs.

[huddlej]

Thank you, @benjaminotter! This looks good to me. I made a couple comments about alternate ways to think about this implementation, but this is good to merge.

[huddlej]

This approach is great and very readable, although you can also use the following idiom for slightly less verbose code:

ignored_characters = distance_map.get("ignored_characters", [])

[huddlej]

This looks good and works for me. One minor consideration is that the tests in the first line will almost always evaluate to True, while the second line will be True much less often. In practice the order of these boolean checks won't make the distance calculations that much slower, but you could consider an approach like this that uses Python's greedy evaluation of logical operators to only check ignored characters when there is a mismatch:

if (node_a_sequences[gene][site] != node_b_sequences[gene][site] and
    node_a_sequences[gene][site] not in ignored_characters and
    node_b_sequences[gene][site] not in ignored_characters):
        # Do some stuff

[benjaminotter]

@huddlej Thank you for the comments, i implemented both improvements in the latest commit.