cleaned up some pandas deprication

deenurp/subcommands/filter_outliers.py

@@ -279,8 +279,7 @@ def parse_usearch_allpairs(filename, seqnames):
     # for each sequence pair, select the longest alignment if there is
     # more than one (chooses first occurrence if there are two the same
     # length).
-    maxidx = data.groupby(['query', 'target']).apply(
-        lambda x: x['align_len'].idxmax())
+    maxidx = data.groupby(['query', 'target'])['align_len'].idxmax()
     data = data.iloc[maxidx]
 
     if set(seqnames) != set(data['query']) | set(data['target']):

deenurp/test/test_outliers.py

@@ -1,3 +1,4 @@
+import io
 import os
 import unittest
 
@@ -95,7 +96,7 @@ def test_choose_clusters(self):
         "medoid":{"0":238.0,"1":null,"2":284.0},
         "dist":{"0":0.0,"1":null,"2":0.089}}"""
 
-        df = pd.read_json(s)
+        df = pd.read_json(io.StringIO(s))
         # output is a set of cluster names (not indices)
         self.assertSetEqual(set(outliers.choose_clusters(df, 2, 0.015)), {0})
         self.assertSetEqual(set(outliers.choose_clusters(df, 2, 0.1)), {0, 1})

deenurp/wrap.py

@@ -9,7 +9,7 @@
 import os.path
 import subprocess
 import re
-from distutils.version import LooseVersion
+from packaging.version import Version
 from io import StringIO
 
 import pandas as pd
@@ -227,10 +227,10 @@ def cmalign_scores(text):
     return pd.read_csv(
         StringIO(text),
         comment="#",
-        delim_whitespace=True,
         dtype=dtypes,
         index_col='seq_name',
-        names=dtypes.keys()
+        names=dtypes.keys(),
+        sep='\s+'
         )
 
 
@@ -281,7 +281,7 @@ def _require_vsearch_version(vsearch=VSEARCH, version=VSEARCH_VERSION):
     vsearch = re.search(r'^vsearch v(?P<vstr>\d+\.\d+\.[^_]+)', output.stderr)
     ver = vsearch.groupdict()['vstr']
 
-    if LooseVersion(ver) < LooseVersion(version):
+    if Version(ver) < Version(version):
         raise MissingDependencyError(
             'vsearch version >= v{} is required, got v{}'.format(version, ver))