class TestSimilarityMatrix(object):
matrix = None
def setUp(self):
self.matrix = SimilarityMatrix('data/similarities.h5')
def tearDown(self):
self.matrix.close()
def test_find_1(self):
result = list(self.matrix.find('2n6i_4FU_frag1', 0.98))
expected = [('2n6i_4FU_frag2', 1.0), ('2n6i_4FU_frag6', 1.0)]
assert_array_almost_equal([r[1] for r in result], [r[1] for r in expected], 3)
eq_([r[0] for r in result], [r[0] for r in result])
def test_iter_first2(self):
myiter = iter(self.matrix)
result = [next(myiter), next(myiter)]
expected = [('2mlm_2W7_frag1', '2mlm_2W7_frag2', 0.5877), ('2mlm_2W7_frag2', '3wvm_STE_frag1', 0.4633)]
assert_array_almost_equal([r[2] for r in result], [r[2] for r in expected], 3)
eq_([(r[0], r[1],) for r in result], [(r[0], r[1],) for r in result])
def test_iter_last(self):
myiter = iter(self.matrix)
result = None
for row in myiter:
result = row
expected = ('3wyl_3KB_frag20', '3wyl_3KB_frag21', 0.999496452277409)
assert_almost_equal(result[2], expected[2], 5)
eq_(result[:2], expected[:2])
def test_simmatrix_import_run_ignore_upper_triangle():
output_fn = tmpname()
tsv = '''frag_id1 frag_id2 score
2mlm_2W7_frag1 2mlm_2W7_frag1 1.0000000000000000
2mlm_2W7_frag2 2mlm_2W7_frag2 1.0000000000000000
2mlm_2W7_frag1 2mlm_2W7_frag2 0.5877164873731594
2mlm_2W7_frag2 3wvm_STE_frag1 0.4633096818493935
2mlm_2W7_frag2 2mlm_2W7_frag1 0.5877164873731594
3wvm_STE_frag1 2mlm_2W7_frag2 0.4633096818493935
'''
inputfile = StringIO(tsv)
try:
script.simmatrix_import_run(inputfile=inputfile,
inputformat='tsv',
simmatrixfn=output_fn,
fragmentsdb='data/fragments.sqlite',
nrrows=2,
ignore_upper_triangle=True)
simmatrix = SimilarityMatrix(output_fn)
result = [r for r in simmatrix]
simmatrix.close()
print(result)
expected = [('2mlm_2W7_frag1', '2mlm_2W7_frag2xx', 0.5877), ('2mlm_2W7_frag2', '3wvm_STE_frag1', 0.4633)]
assert_array_almost_equal([r[2] for r in result], [r[2] for r in expected], 3)
assert [(r[0], r[1],) for r in result] == [(r[0], r[1],) for r in result]
finally:
if os.path.exists(output_fn):
os.remove(output_fn)
def test_simmatrix_filter_keep(keep_fragments_db, output_simmatrix_fn):
script.simmatrix_filter('data/similarities.h5', output_simmatrix_fn, keep_fragments_db, None)
result = SimilarityMatrix(output_simmatrix_fn)
nr_pairs = len(result.pairs)
result.close()
assert nr_pairs == 93
def test_simmatrix_import_run_ignore_upper_triangle():
output_fn = tmpname()
tsv = '''frag_id1 frag_id2 score
2mlm_2W7_frag1 2mlm_2W7_frag1 1.0000000000000000
2mlm_2W7_frag2 2mlm_2W7_frag2 1.0000000000000000
2mlm_2W7_frag1 2mlm_2W7_frag2 0.5877164873731594
2mlm_2W7_frag2 3wvm_STE_frag1 0.4633096818493935
2mlm_2W7_frag2 2mlm_2W7_frag1 0.5877164873731594
3wvm_STE_frag1 2mlm_2W7_frag2 0.4633096818493935
'''
inputfile = StringIO(tsv)
try:
script.simmatrix_import_run(inputfile=inputfile,
inputformat='tsv',
simmatrixfn=output_fn,
fragmentsdb='data/fragments.sqlite',
nrrows=2,
ignore_upper_triangle=True)
simmatrix = SimilarityMatrix(output_fn)
result = [r for r in simmatrix]
simmatrix.close()
print(result)
expected = [('2mlm_2W7_frag1', '2mlm_2W7_frag2xx', 0.5877), ('2mlm_2W7_frag2', '3wvm_STE_frag1', 0.4633)]
assert_array_almost_equal([r[2] for r in result], [r[2] for r in expected], 3)
assert [(r[0], r[1],) for r in result] == [(r[0], r[1],) for r in result]
finally:
if os.path.exists(output_fn):
os.remove(output_fn)
def test_simmatrix_importfpneigh_run_ignore_upper_triangle():
output_fn = tmpname()
tsv = '''Compounds similar to 2mlm_2W7_frag1:
2mlm_2W7_frag1 1.0000
2mlm_2W7_frag2 0.5877
Compounds similar to 2mlm_2W7_frag2:
2mlm_2W7_frag2 1.0000
2mlm_2W7_frag1 0.5877
3wvm_STE_frag1 0.4633
'''
inputfile = StringIO(tsv)
try:
script.simmatrix_importfpneigh_run(inputfile=inputfile,
simmatrixfn=output_fn,
fragmentsdb='data/fragments.sqlite',
nrrows=3,
ignore_upper_triangle=True)
simmatrix = SimilarityMatrix(output_fn)
rows = [r for r in simmatrix]
simmatrix.close()
expected = [(u'2mlm_2W7_frag1', u'2mlm_2W7_frag2', 0.5877), (u'2mlm_2W7_frag2', u'3wvm_STE_frag1', 0.4633)]
assert rows == expected
finally:
os.remove(output_fn)
def test_simmatrix_filter_keep(keep_fragments_db, output_simmatrix_fn):
script.simmatrix_filter('data/similarities.h5', output_simmatrix_fn, keep_fragments_db, None)
result = SimilarityMatrix(output_simmatrix_fn)
nr_pairs = len(result.pairs)
result.close()
assert nr_pairs == 93
def test_simmatrix_importfpneigh_run_ignore_upper_triangle():
output_fn = tmpname()
tsv = '''Compounds similar to 2mlm_2W7_frag1:
2mlm_2W7_frag1 1.0000
2mlm_2W7_frag2 0.5877
Compounds similar to 2mlm_2W7_frag2:
2mlm_2W7_frag2 1.0000
2mlm_2W7_frag1 0.5877
3wvm_STE_frag1 0.4633
'''
inputfile = StringIO(tsv)
try:
script.simmatrix_importfpneigh_run(inputfile=inputfile,
simmatrixfn=output_fn,
fragmentsdb='data/fragments.sqlite',
nrrows=3,
ignore_upper_triangle=True)
simmatrix = SimilarityMatrix(output_fn)
rows = [r for r in simmatrix]
simmatrix.close()
expected = [(u'2mlm_2W7_frag1', u'2mlm_2W7_frag2', 0.5877), (u'2mlm_2W7_frag2', u'3wvm_STE_frag1', 0.4633)]
assert rows == expected
finally:
os.remove(output_fn)
class TestWebservice(object):
def setUp(self):
self.matrix = SimilarityMatrix('data/similarities.h5')
self.app = server.wsgi_app(self.matrix)
def tearDown(self):
self.matrix.close()
def test_get_similar_fragments(self):
fragment_id = '3j7u_NDP_frag24'
cutoff = 0.85
with self.app.app.test_request_context():
result = server.get_similar_fragments(fragment_id, cutoff, 1000)
expected = [
{'query_frag_id': '3j7u_NDP_frag24', 'hit_frag_id': '3j7u_NDP_frag23', 'score': 0.8991},
]
eq_(result, expected)
def test_get_version(self):
result = server.get_version()
expected = {'version': __version__}
eq_(result, expected)
def test_wsgi_app(self):
eq_(self.app.app.config['matrix'], self.matrix)
def test_simmatrix_filter_skip(output_simmatrix_fn):
labels2skip = StringIO('2mm3_CHO_frag1\n')
script.simmatrix_filter('data/similarities.h5', output_simmatrix_fn, None, labels2skip)
result = SimilarityMatrix(output_simmatrix_fn)
nr_pairs = len(result.pairs)
result.close()
assert nr_pairs == 11857
def test_simmatrix_filter_skip(output_simmatrix_fn):
labels2skip = StringIO('2mm3_CHO_frag1\n')
script.simmatrix_filter('data/similarities.h5', output_simmatrix_fn, None, labels2skip)
result = SimilarityMatrix(output_simmatrix_fn)
nr_pairs = len(result.pairs)
result.close()
assert nr_pairs == 11857
def test_merge():
infiles = [tmpname(), tmpname(), tmpname()]
outfile = tmpname()
try:
# fill infiles
inmatrix1 = SimilarityMatrix(infiles[0], 'w', 1, 2**16-1, 2)
inmatrix1.update([('a', 'b', 0.2)], {'a': 1, 'b': 2, 'c': 3})
inmatrix1.close()
# matrix with same labels -> copy pairs table by dump/append, ignores labels tables
inmatrix2 = SimilarityMatrix(infiles[1], 'w', 2, 2**16-1, 3)
inmatrix2.update([('a', 'c', 0.6)], {'a': 1, 'b': 2, 'c': 3})
inmatrix2.close()
# matrix generated with different labels -> copy pairs table by iterate/update, adds missing labels
inmatrix3 = SimilarityMatrix(infiles[2], 'w', 2, 2**16-1, 3)
inmatrix3.update([('b', 'e', 0.4), ('e', 'f', 0.8)], {'b': 1, 'e': 2, 'f': 3})
inmatrix3.close()
pairs.merge(infiles, outfile)
# compare it
outmatrix = SimilarityMatrix(outfile)
result = list(outmatrix)
outmatrix.close()
expected = [('a', 'b', 0.2), ('a', 'c', 0.6), ('b', 'e', 0.4), ('e', 'f', 0.8)]
assert result == expected
finally:
for infile in infiles:
if os.path.isfile(infile):
os.remove(infile)
if os.path.isfile(outfile):
os.remove(outfile)
class SimilarityMatrixInMemory(object):
def __init__(self):
self.matrix_fn = tmpname()
self.matrix = SimilarityMatrix(self.matrix_fn, 'a', driver='H5FD_CORE', driver_core_backing_store=0)
def __enter__(self):
return self.matrix
def __exit__(self, exc_type, exc_val, exc_tb):
self.close()
def close(self):
self.matrix.close()
if os.path.isfile(self.matrix_fn):
os.remove(self.matrix_fn)
class SimilarityMatrixInMemory(object):
def __init__(self):
self.matrix_fn = tmpname()
self.matrix = SimilarityMatrix(self.matrix_fn,
'a',
driver='H5FD_CORE',
driver_core_backing_store=0)
def __enter__(self):
return self.matrix
def __exit__(self, exc_type, exc_val, exc_tb):
self.close()
def close(self):
self.matrix.close()
if os.path.isfile(self.matrix_fn):
os.remove(self.matrix_fn)
def test_merge():
infiles = [tmpname(), tmpname(), tmpname()]
outfile = tmpname()
try:
# fill infiles
inmatrix1 = SimilarityMatrix(infiles[0], 'w', 1, 2**16 - 1, 2)
inmatrix1.update([('a', 'b', 0.2)], {'a': 1, 'b': 2, 'c': 3})
inmatrix1.close()
# matrix with same labels -> copy pairs table by dump/append, ignores labels tables
inmatrix2 = SimilarityMatrix(infiles[1], 'w', 2, 2**16 - 1, 3)
inmatrix2.update([('a', 'c', 0.6)], {'a': 1, 'b': 2, 'c': 3})
inmatrix2.close()
# matrix generated with different labels -> copy pairs table by iterate/update, adds missing labels
inmatrix3 = SimilarityMatrix(infiles[2], 'w', 2, 2**16 - 1, 3)
inmatrix3.update([('b', 'e', 0.4), ('e', 'f', 0.8)], {
'b': 1,
'e': 2,
'f': 3
})
inmatrix3.close()
pairs.merge(infiles, outfile)
# compare it
outmatrix = SimilarityMatrix(outfile)
result = list(outmatrix)
outmatrix.close()
expected = [('a', 'b', 0.2), ('a', 'c', 0.6), ('b', 'e', 0.4),
('e', 'f', 0.8)]
assert result == expected
finally:
for infile in infiles:
if os.path.isfile(infile):
os.remove(infile)
if os.path.isfile(outfile):
os.remove(outfile)
class TestFrozenSimilarityMatrix(object):
pair_matrix_fn = None
pair_matrix = None
matrix_fn = None
matrix = None
def setUp(self):
self.pair_matrix_fn = tmpname()
self.pair_matrix = SimilarityMatrix(self.pair_matrix_fn, "a", driver="H5FD_CORE", driver_core_backing_store=0)
labels = {"a": 0, "b": 1, "c": 2, "d": 3}
similarities = [("a", "b", 0.9), ("a", "c", 0.5), ("b", "c", 0.6), ("d", "c", 0.7)]
self.pair_matrix.update(similarities, labels)
self.matrix_fn = tmpname()
self.matrix = FrozenSimilarityMatrix(self.matrix_fn, "a", driver="H5FD_CORE", driver_core_backing_store=0)
def tearDown(self):
self.pair_matrix.close()
if os.path.isfile(self.pair_matrix_fn):
os.remove(self.pair_matrix_fn)
self.matrix.close()
if os.path.isfile(self.matrix_fn):
os.remove(self.matrix_fn)
def test_from_pairs_defaults(self):
self.matrix.from_pairs(self.pair_matrix, 10)
result = self.matrix.to_pandas()
labels = ["a", "b", "c", "d"]
expected = pd.DataFrame(
[[0.0, 0.9, 0.5, 0.0], [0.9, 0.0, 0.6, 0.0], [0.5, 0.6, 0.0, 0.7], [0.0, 0.0, 0.7, 0.0]],
index=labels,
columns=labels,
)
pdt.assert_almost_equal(result, expected)
def test_from_pairs_multiframe(self):
self.matrix.from_pairs(self.pair_matrix, 1, None, False)
result = self.matrix.to_pandas()
labels = ["a", "b", "c", "d"]
expected = pd.DataFrame(
[[0.0, 0.9, 0.5, 0.0], [0.9, 0.0, 0.6, 0.0], [0.5, 0.6, 0.0, 0.7], [0.0, 0.0, 0.7, 0.0]],
index=labels,
columns=labels,
)
pdt.assert_almost_equal(result, expected)
def test_from_pairs_limited(self):
self.matrix.from_pairs(self.pair_matrix, 1, 1, False)
result = self.matrix.to_pandas()
labels = ["a", "b", "c", "d"]
expected = pd.DataFrame(
[[0.0, 0.9, 0.0, 0.0], [0.9, 0.0, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0]],
index=labels,
columns=labels,
)
pdt.assert_almost_equal(result, expected)
def test_from_pairs_singlesided(self):
self.matrix.from_pairs(self.pair_matrix, 10, None, True)
result = self.matrix.to_pandas()
print(result)
labels = ["a", "b", "c", "d"]
expected = pd.DataFrame(
[[0.0, 0.9, 0.5, 0.0], [0.0, 0.0, 0.6, 0.0], [0.0, 0.0, 0.0, 0.0], [0.0, 0.0, 0.7, 0.0]],
index=labels,
columns=labels,
)
pdt.assert_almost_equal(result, expected)
def test_find_defaults(self):
self.matrix.from_pairs(self.pair_matrix, 10)
hits = self.matrix.find("c", 0.55)
expected = [("d", 0.7), ("b", 0.6)]
eq_(hits, expected)
def test_find_limit(self):
self.matrix.from_pairs(self.pair_matrix, 10)
hits = self.matrix.find("c", 0.55, 1)
expected = [("d", 0.7)]
eq_(hits, expected)
def test_find_cutoffhigh_nohits(self):
self.matrix.from_pairs(self.pair_matrix, 10)
hits = self.matrix.find("c", 0.9)
expected = []
eq_(hits, expected)
def test_find_badkey_keyerror(self):
self.matrix.from_pairs(self.pair_matrix, 10)
with assert_raises(KeyError):
self.matrix.find("f", 0.45)
def test_find_singlesided(self):
self.matrix.from_pairs(self.pair_matrix, 10, None, True)
print(self.matrix.scores.read())
hits = self.matrix.find("c", 0.0)
expected = []
eq_(hits, expected)
def matrix():
sim_matrix = SimilarityMatrix('data/similarities.h5')
yield sim_matrix
sim_matrix.close();