🚧 filter: Exclude sequences with ambiguous month

workaround for an augur filter bug that doesn't treat XXXX-XX-XX as a
"missing date" for date filters (FIXME: write up the bug)

Author: victorlin

phylogenetic/defaults/config.yaml

@@ -34,6 +34,7 @@ builds:
       exclude: defaults/exclude.txt
       include: defaults/drc-2025/include.txt
       min_date: "2025-08-01"
+      exclude_ambiguous_dates_by: month
     refine:
       timetree: false
       coalescent: skyline

phylogenetic/rules/prepare_sequences.smk

@@ -57,6 +57,7 @@ rule filter:
         min_length = lambda w: conditional("--min-length", config["builds"][w.build]["filter"].get("min_length")),
         min_date = lambda w: conditional("--min-date", config["builds"][w.build]["filter"].get("min_date")),
         max_date = lambda w: conditional("--max-date", config["builds"][w.build]["filter"].get("max_date")),
+        exclude_ambiguous_dates_by = lambda w: conditional("--exclude-ambiguous-dates-by", config["builds"][w.build]["filter"].get("exclude_ambiguous_dates_by")),
         exclude_where = lambda w: conditional("--exclude-where", config["builds"][w.build]["filter"].get("exclude_where")),
         group_by = lambda w: conditional("--group-by", config["builds"][w.build]["filter"].get("group_by")),
         subsample_max_sequences = lambda w: conditional("--subsample-max-sequences", config["builds"][w.build]["filter"].get("subsample_max_sequences")),
@@ -75,6 +76,7 @@ rule filter:
             {params.min_length:q} \
             {params.min_date:q} \
             {params.max_date:q} \
+            {params.exclude_ambiguous_dates_by:q} \
             {params.exclude_where:q} \
             {params.group_by:q} \
             {params.subsample_max_sequences:q} \