def test_correlation_linkage_1(self):
table = pd.DataFrame([[1, 1, 0, 0, 0], [0, 1, 1, 0, 0],
[0, 0, 1, 1, 0], [0, 0, 0, 1, 1]],
columns=['s1', 's2', 's3', 's4', 's5'],
index=['o1', 'o2', 'o3', 'o4']).T
exp_str = ('((o1:0.574990173931,o2:0.574990173931)y1:0.773481312844,'
'(o3:0.574990173931,o4:0.574990173931)y2:0.773481312844)'
'y0;\n')
res_tree = correlation_linkage(table + 0.1)
self.assertEqual(exp_str, str(res_tree))
def test_correlation_linkage_2(self):
t = pd.DataFrame([[1, 1, 2, 3, 1, 4], [2, 2, 0.1, 4, 1, .1],
[3, 3.1, 2, 3, 2, 2], [4.1, 4, 0.2, 1, 1, 2.5]],
index=['S1', 'S2', 'S3', 'S4'],
columns=['F1', 'F2', 'F3', 'F4', 'F5', 'F6'])
exp_str = ('((F4:0.228723591874,(F5:0.074748541601,'
'(F1:0.00010428164962,F2:0.00010428164962)'
'y4:0.0746442599513)y3:0.153975050273)'
'y1:0.70266138894,(F3:0.266841737789,F6:0.266841737789)'
'y2:0.664543243026)y0;\n')
res_tree = correlation_linkage(t)
self.assertEqual(exp_str, str(res_tree))
def test_correlation_linkage_1(self):
table = pd.DataFrame([[1, 1, 0, 0, 0], [0, 1, 1, 0, 0],
[0, 0, 1, 1, 0], [0, 0, 0, 1, 1]],
columns=['s1', 's2', 's3', 's4', 's5'],
index=['o1', 'o2', 'o3', 'o4']).T
exp_str = ('((o1:0.574990173931,o2:0.574990173931)y1:0.773481312844,'
'(o3:0.574990173931,o4:0.574990173931)y2:0.773481312844)'
'y0;\n')
exp_tree = TreeNode.read([exp_str])
res_tree = correlation_linkage(table + 0.1)
# only check for tree topology since checking for floating point
# numbers on the branches is still tricky.
self.assertEqual(exp_tree.ascii_art(), res_tree.ascii_art())
def test_correlation_linkage_2(self):
t = pd.DataFrame([[1, 1, 2, 3, 1, 4],
[2, 2, 0.1, 4, 1, .1],
[3, 3.1, 2, 3, 2, 2],
[4.1, 4, 0.2, 1, 1, 2.5]],
index=['S1', 'S2', 'S3', 'S4'],
columns=['F1', 'F2', 'F3', 'F4', 'F5', 'F6'])
exp_str = ('((F4:0.228723591874,(F5:0.074748541601,'
'(F1:0.00010428164962,F2:0.00010428164962)'
'y4:0.0746442599513)y3:0.153975050273)'
'y1:0.70266138894,(F3:0.266841737789,F6:0.266841737789)'
'y2:0.664543243026)y0;\n')
exp_tree = TreeNode.read([exp_str])
res_tree = correlation_linkage(t)
self.assertEqual(exp_tree.ascii_art(), res_tree.ascii_art())
def correlation_clustering(table: pd.DataFrame) -> skbio.TreeNode:
""" Builds a tree for features based on correlation.
Parameters
----------
table : pd.DataFrame
Contingency table where rows are samples and columns are features.
In addition, the table must have strictly nonzero values.
Returns
-------
skbio.TreeNode
Represents the partitioning of features with respect to correlation.
"""
t = correlation_linkage(table)
return t
def test_correlation_linkage_1(self):
table = pd.DataFrame(
[[1, 1, 0, 0, 0],
[0, 1, 1, 0, 0],
[0, 0, 1, 1, 0],
[0, 0, 0, 1, 1]],
columns=['s1', 's2', 's3', 's4', 's5'],
index=['o1', 'o2', 'o3', 'o4']).T
exp_str = ('((o1:0.574990173931,o2:0.574990173931)y1:0.773481312844,'
'(o3:0.574990173931,o4:0.574990173931)y2:0.773481312844)'
'y0;\n')
exp_tree = TreeNode.read([exp_str])
res_tree = correlation_linkage(table+0.1)
# only check for tree topology since checking for floating point
# numbers on the branches is still tricky.
self.assertEqual(exp_tree.ascii_art(), res_tree.ascii_art())