Changes intended to assist with investigation of issue #28 (possibly also issue #23)

cluster/matrix.py

@@ -22,6 +22,30 @@
 
 logger = logging.getLogger(__name__)
 
+def _encapsulate_item_for_combinfunc(item):
+    """
+    This function has been extracted in order to 
+    make Github issue #28 easier to investigate.
+    It replaces the following two lines of code, 
+    which occur twice in method genmatrix, just
+    before the invocation of combinfunc.
+        if not hasattr(item, '__iter__') or isinstance(item, tuple):
+            item = [item]
+    Logging has been added to the original two lines
+    and shows that the behaviour of this snippet
+    has changed between Python2.7 and Python3.5.
+    """
+    encapsulated_item = None
+    if not hasattr(item, '__iter__') or isinstance(item, tuple):
+        encapsulated_item = [item]
+    else:
+        encapsulated_item = item
+    logging.debug(
+        "item class:%s encapsulated as:%s ",
+        item.__class__, encapsulated_item.__class__
+    )
+    return encapsulated_item
+
 
 class Matrix(object):
     """
@@ -123,10 +147,17 @@ def genmatrix(self, num_processes=1):
                         num_tasks_completed += 1
                 else:
                     # Otherwise do it here, in line
+                    """
                     if not hasattr(item, '__iter__') or isinstance(item, tuple):
                         item = [item]
                     if not hasattr(item2, '__iter__') or isinstance(item2, tuple):
                         item2 = [item2]
+                    """
+                    # See the comment in function _encapsulate_item_for_combinfunc
+                    # for details of why the lines above have been replaced
+                    # by function invocations
+                    item = _encapsulate_item_for_combinfunc(item)
+                    item2 = _encapsulate_item_for_combinfunc(item2)
                     row[col_index] = self.combinfunc(item, item2)
 
             if self.symmetric:

cluster/test/test_hierarchical.py

@@ -231,12 +231,42 @@ def euclidian_distance(a, b):
         result = cl.getlevel(40)
         self.assertIsNotNone(result)
 
+class Issue28TestCase(Py23TestCase):
+    '''
+    Test case to cover the case where the data consist
+    of dictionary keys, and the distance function executes 
+    on the values these keys are associated with in the
+    dictionary, rather than the keys themselves.
+
+    Behaviour for this test case differs between Python2.7
+    and Python3.5: on 2.7 the test behaves as expected, 
+
+    See Github issue #28.
+    '''
+
+    def testIssue28(self):
+        "Issue28 (Hierarchical Clustering)"
 
+        points1D = {
+            'p4' : 5, 'p2' : 6, 'p7' : 10,
+            'p9' : 120, 'p10' : 121, 'p11' : 119,
+        }
+
+        distance_func = lambda a,b : abs(points1D[a]-points1D[b])
+        cl = HierarchicalClustering(list(points1D.keys()), distance_func)
+        result = cl.getlevel(20)
+        self.assertIsNotNone(result)
+
 if __name__ == '__main__':
+
+    import logging
+
     suite = unittest.TestSuite((
         unittest.makeSuite(HClusterIntegerTestCase),
         unittest.makeSuite(HClusterSmallListTestCase),
         unittest.makeSuite(HClusterStringTestCase),
+        unittest.makeSuite(Issue28TestCase),
     ))
 
+    logging.basicConfig(level=logging.DEBUG)
     unittest.TextTestRunner(verbosity=2).run(suite)

cluster/test/test_linkage.py

@@ -29,3 +29,14 @@ def test_average_distance(self):
         result = average(self.set_a, self.set_b, self.dist)
         expected = 22.5
         self.assertEqual(result, expected)
+
+if __name__ == '__main__':
+
+    import logging
+
+    suite = unittest.TestSuite((
+        unittest.makeSuite(LinkageMethods),
+    ))
+
+    logging.basicConfig(level=logging.DEBUG)
+    unittest.TextTestRunner(verbosity=2).run(suite)