Merge pull request #29 from sjanssen2/py35

Py35

.travis.yml

@@ -1,7 +1,7 @@
 # Modified from https://github.com/biocore/biom-format/
 language: python
 env:
-  - PYTHON_VERSION=2.7
+  - PYTHON_VERSION=3.6
 before_install:
   - wget https://repo.continuum.io/miniconda/Miniconda2-latest-Linux-x86_64.sh -O miniconda.sh
   - export MINICONDA_PREFIX="$HOME/miniconda"
@@ -14,7 +14,7 @@ before_install:
 install:
   - conda create --yes -n env_name python=$PYTHON_VERSION pip numpy scipy matplotlib nose pep8 Sphinx
   - source activate env_name
-  - pip install future coveralls natsort pyflakes flake8 python-dateutil scikit-bio==0.4.2 click
+  - pip install future coveralls natsort pyflakes flake8 python-dateutil scikit-bio click
   - pip install -e . --no-deps
 script:
   - if [ ${WITH_DOCTEST} ]; then PYTHONWARNINGS=ignore nosetests --with-doctest --with-coverage -I DONOTIGNOREANYTHING; else PYTHONWARNINGS=ignore nosetests --with-coverage -I DONOTIGNOREANYTHING; fi

t2t/consensus.py

@@ -6,7 +6,7 @@
 make_consensus_tree
 etc...
 """
-from nlevel import RANK_ORDER
+from t2t.nlevel import RANK_ORDER
 from numpy import zeros, where, logical_or, long
 
 
@@ -25,7 +25,7 @@ def taxa_score(master, reps):
     for rep in reps:
         n_reps += 1
 
-        for id_, con in rep.iteritems():
+        for id_, con in rep.items():
             if id_ not in master_ids:
                 raise KeyError("Unknown key %s in replicate" % id_)
 
@@ -49,7 +49,7 @@ def taxa_score(master, reps):
 
 def merge_taxa_strings_and_scores(master, scores):
     """Merge taxa strings and their scores, return {id_:(taxa,score)}"""
-    return {k: zip(v, scores[k]) for k, v in master.items()}
+    return {k: list(zip(v, scores[k])) for k, v in master.items()}
 
 
 def taxa_score_hash(master, reps):
@@ -108,7 +108,12 @@ def get_consensus_stats(consensus_map):
     cons = consensus_map.values()
     total_cons = len(cons)
 
-    rank_names = [c[0] for c in cons[0]]
+    rank_names = None
+    for co in cons:
+        if co is None:
+            continue
+        rank_names = [c[0] for c in co if c is not None]
+        break
     for idx, rank in enumerate(rank_names):
         # collect all cons that are classified (ie more info than k__)
         klassed = [c[idx].lower()

t2t/consistency.py

@@ -57,7 +57,7 @@ def calculate(self, tree, rooted):
         consistency_index = {i: defaultdict(int) for i in range(self.n_ranks)}
         for n in tree.traverse(include_self=True):
             for rank in range(self.n_ranks):
-                for name, total_taxa_cnt in self.taxa_counts[rank].iteritems():
+                for name, total_taxa_cnt in self.taxa_counts[rank].items():
                     node_taxa_count = n.TaxaCount[rank].get(name, 0)
                     incongruent_taxa = n.NumTipsRank[rank] - node_taxa_count
 
@@ -93,7 +93,7 @@ def write_taxon_consistency(self, output_file, consistency_index):
         fout = open(output_file, 'w')
         fout.write('Taxon\tCount\tConsistency\n')
         for rank in range(self.n_ranks):
-            for name, consistency in consistency_index[rank].iteritems():
+            for name, consistency in consistency_index[rank].items():
                 fout.write('%s\t%d\t%.3f\n' % (name,
                                                self.taxa_counts[rank][name],
                                                consistency))
@@ -116,7 +116,7 @@ def write_rank_consistency(self, output_file, consistency_index, min_taxa,
         fout.write('Rank #\tRank prefix\t# taxon\tAverage consistency\n')
         for rank in range(self.n_ranks):
             val = []
-            for name, consistency in consistency_index[rank].iteritems():
+            for name, consistency in consistency_index[rank].items():
                 if self.taxa_counts[rank][name] >= min_taxa:
                     val.append(consistency)
 

t2t/nlevel.py

@@ -6,8 +6,6 @@
 from skbio import TreeNode
 from t2t.util import unzip
 import re
-import os
-import io
 
 __author__ = "Daniel McDonald"
 __copyright__ = "Copyright 2011, The tax2tree project"
@@ -517,6 +515,7 @@ def name_node_score_fold(tree, score_f=fmeasure, tiebreak_f=min_tips,
 
     return used_scores
 
+
 def score_tree(tree, verbose=False):
     """Scores the tree based on RankNameScores and tip coverage
 
@@ -572,6 +571,7 @@ def make_consensus_tree(cons_split, check_for_rank=True, tips=None):
     god_node.Rank = None
 
     base = list(cons_split)[0]
+    #  SMJ base = next(iter(cons_split))
     cur_node = god_node
 
     # create a base path in the tree

t2t/remap.py

@@ -4,8 +4,8 @@
 def parse_otu_map(lines):
     """Returns {rep: [members]}, members include rep"""
     res = {}
-    for l in lines:
-        fields = l.strip().split('\t')
+    for line in lines:
+        fields = line.strip().split('\t')
         rep = fields[1]
         members = fields[1:]
         res[rep] = members
@@ -15,7 +15,7 @@ def parse_otu_map(lines):
 def members_to_rep(otus):
     """Provides a lookup for a cluster member to its rep"""
     res = {}
-    for rep, members in otus.iteritems():
+    for rep, members in otus.items():
         for member in members:
             res[member] = rep
     return res
@@ -26,7 +26,7 @@ def remap_taxonomy(mapping, taxa):
     res = {}
     reps = members_to_rep(mapping)
 
-    for tax_id, tax_str in taxa.iteritems():
+    for tax_id, tax_str in taxa.items():
         rep = reps.get(tax_id, None)
 
         if rep is None:

tests/test_consensus.py

@@ -145,20 +145,20 @@ def test_hash_cons(self):
             'c': [None, None, None, None, None, None, None],
             'd': ['k__k2', 'p__x', 'c__c1', 'o__o10', 'f__f2', 'g__g2',
                   's__s3']}
-        exp = array([map(hash, input['a']),
-                     map(hash, input['d']),
-                     map(hash, input['c']),
-                     map(hash, input['b'])])
+        exp = array([*map(hash, input['a']),
+                     *map(hash, input['d']),
+                     *map(hash, input['c']),
+                     *map(hash, input['b'])])
         obs = hash_cons(input, ['a', 'd', 'c', 'b'], 7)
-        self.assertTrue(array_equal(obs, exp))
+        self.assertTrue(array_equal(obs.flatten(), exp))
 
-        exp = array([map(hash, input['a']),
-                     map(hash, input['d']),
-                     [0, 0, 0, 0, 0, 0, 0],
-                     map(hash, input['c']),
-                     map(hash, input['b'])])
+        exp = array([*map(hash, input['a']),
+                     *map(hash, input['d']),
+                     *map(hash, [0, 0, 0, 0, 0, 0, 0]),
+                     *map(hash, input['c']),
+                     *map(hash, input['b'])])
         obs = hash_cons(input, ['a', 'd', 'e', 'c', 'b'], 7)
-        self.assertTrue(array_equal(obs, exp))
+        self.assertTrue(array_equal(obs.flatten(), exp))
 
     def test_get_consensus_stats(self):
         """Produces the correct stats"""
@@ -176,7 +176,7 @@ def test_get_consensus_stats(self):
                      'f': set(['f__f1', 'f__f2']), 'g': set(['g__g1',
                                                              'g__g2']),
                      's': set(['s__s1', 's__s2', 's__s3'])}
-        #exp_names = {'k':2,'p':1,'c':2,'o':1,'f':2,'g':2,'s':3}
+        #  exp_names = {'k':2,'p':1,'c':2,'o':1,'f':2,'g':2,'s':3}
         obs_nseqs, obs_names = get_consensus_stats(input)
         self.assertEqual(obs_nseqs, exp_nseqs)
         self.assertEqual(obs_names, exp_names)