def test_mh_asymmetric_merge(track_abundance):
# test merging two asymmetric (different size) MHs
a = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 40, 2):
a.add_hash(i)
# different size: 10
b = MinHash(10, 10, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
d = b.merge(a)
assert len(a) == 20
assert len(b) == 10
assert len(c) == len(a)
assert len(d) == len(b)
# can't compare different sizes without downsampling
with pytest.raises(TypeError):
d.compare(a)
a = a.downsample_n(d.num)
print(a.get_mins())
print(d.get_mins())
assert d.compare(a) == 1.0
c = c.downsample_n(b.num)
assert c.compare(b) == 1.0
def test_mh_asymmetric_merge(track_abundance):
# test merging two asymmetric (different size) MHs
a = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 40, 2):
a.add_hash(i)
# different size: 10
b = MinHash(10, 10, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
d = b.merge(a)
assert len(a) == 20
assert len(b) == 10
assert len(c) == len(a)
assert len(d) == len(b)
# can't compare different sizes without downsampling
with pytest.raises(TypeError):
d.compare(a)
a = a.downsample_n(d.num)
print(a.get_mins())
print(d.get_mins())
assert d.compare(a) == 1.0
c = c.downsample_n(b.num)
assert c.compare(b) == 1.0
def test_minhash_abund_merge_flat_2():
# this targets a segfault caused by trying to merge
# a signature with abundance and a signature without abundance.
a = MinHash(0, 10, track_abundance=True, max_hash=5000)
b = MinHash(0, 10, max_hash=5000)
for i in range(0, 10, 2):
a.add_hash(i)
for j in range(0, 10, 3):
b.add_hash(i)
a.merge(b)
def test_minhash_abund_merge_flat_2():
# this targets a segfault caused by trying to merge
# a signature with abundance and a signature without abundance.
a = MinHash(0, 10, track_abundance=True, max_hash=5000)
b = MinHash(0, 10, max_hash=5000)
for i in range(0, 10, 2):
a.add_hash(i)
for j in range(0, 10, 3):
b.add_hash(i)
a.merge(b)
def test_mh_merge_empty_num(track_abundance):
# test merging two identically configured minhashes, one empty
a = MinHash(20, 10, track_abundance=track_abundance)
b = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
d = b.merge(a)
assert len(c)
assert len(c) == len(d)
assert c.get_mins() == d.get_mins()
assert round(c.similarity(d), 3) == 1.0
assert round(d.similarity(c), 3) == 1.0
def test_mh_merge_empty_scaled(track_abundance):
# test merging two identically configured minhashes, one empty
a = MinHash(0, 10, scaled=1, track_abundance=track_abundance)
b = MinHash(0, 10, scaled=1, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
d = b.merge(a)
assert len(c)
assert len(c) == len(d)
assert c.get_mins() == d.get_mins()
assert c.compare(d) == 1.0
assert d.compare(c) == 1.0
def test_mh_merge(track_abundance):
# test merging two identically configured minhashes
a = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 40, 2):
a.add_hash(i)
b = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
d = b.merge(a)
assert len(c) == len(d)
assert c.get_mins() == d.get_mins()
assert c.compare(d) == 1.0
assert d.compare(c) == 1.0
def test_mh_merge(track_abundance):
# test merging two identically configured minhashes
a = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 40, 2):
a.add_hash(i)
b = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
d = b.merge(a)
assert len(c) == len(d)
assert c.get_mins() == d.get_mins()
assert c.compare(d) == 1.0
assert d.compare(c) == 1.0
def test_mh_merge_check_length(track_abundance):
a = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 40, 2):
a.add_hash(i)
b = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
assert len(c.get_mins()) == 20
def test_mh_merge_check_length(track_abundance):
a = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 40, 2):
a.add_hash(i)
b = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
assert(len(c.get_mins()) == 20)
def test_mh_merge(track_abundance):
# test merging two identically configured minhashes
a = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 40, 2):
a.add_hash(i)
b = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
d = b.merge(a)
assert len(c) == len(d)
assert c.get_mins() == d.get_mins()
if track_abundance:
assert round(c.similarity(d), 3) == 0.91
assert round(d.similarity(c), 3) == 0.91
else:
assert round(c.similarity(d), 3) == 1.0
assert round(d.similarity(c), 3) == 1.0
def test_mh_merge_check_length2(track_abundance):
# merged MH doesn't have full number of elements
a = MinHash(4, 10, track_abundance=track_abundance)
a.add_hash(3)
a.add_hash(1)
a.add_hash(4)
b = MinHash(4, 10, track_abundance=track_abundance)
b.add_hash(3)
b.add_hash(1)
b.add_hash(4)
c = a.merge(b)
assert len(c.get_mins()) == 3
def test_mh_merge_check_length2(track_abundance):
# merged MH doesn't have full number of elements
a = MinHash(4, 10, track_abundance=track_abundance)
a.add_hash(3)
a.add_hash(1)
a.add_hash(4)
b = MinHash(4, 10, track_abundance=track_abundance)
b.add_hash(3)
b.add_hash(1)
b.add_hash(4)
c = a.merge(b)
assert(len(c.get_mins()) == 3)
def test_mh_asymmetric_merge(track_abundance):
# test merging two asymmetric (different size) MHs
a = MinHash(20, 10, track_abundance=track_abundance)
for i in range(0, 40, 2):
a.add_hash(i)
# different size: 10
b = MinHash(10, 10, track_abundance=track_abundance)
for i in range(0, 80, 4):
b.add_hash(i)
c = a.merge(b)
d = b.merge(a)
assert len(a) == 20
assert len(b) == 10
assert len(c) == len(a)
assert len(d) == len(b)
# can't use jaccard on different nums without downsampling
with pytest.raises(TypeError):
d.jaccard(a)
a = a.downsample_n(d.num)
print(a.get_mins())
print(d.get_mins())
if track_abundance:
assert round(d.similarity(a), 3) == 0.91
else:
assert round(d.similarity(a), 3) == 1.0
c = c.downsample_n(b.num)
if track_abundance:
assert round(c.similarity(b), 3) == 0.91
else:
assert c.similarity(b) == 1.0
def test_mh_merge_diff_ksize(track_abundance):
a = MinHash(20, 5, track_abundance=track_abundance)
b = MinHash(20, 6, track_abundance=track_abundance)
with pytest.raises(ValueError):
a.merge(b)
def test_mh_merge_diff_protein(track_abundance):
a = MinHash(20, 5, False, track_abundance=track_abundance)
b = MinHash(20, 5, True, track_abundance=track_abundance)
with pytest.raises(ValueError):
a.merge(b)
def test_mh_merge_diff_protein(track_abundance):
a = MinHash(20, 5, False, track_abundance=track_abundance)
b = MinHash(20, 5, True, track_abundance=track_abundance)
with pytest.raises(ValueError):
a.merge(b)
def test_mh_merge_diff_ksize(track_abundance):
a = MinHash(20, 5, track_abundance=track_abundance)
b = MinHash(20, 6, track_abundance=track_abundance)
with pytest.raises(ValueError):
a.merge(b)
def test_mh_merge_typeerror(track_abundance):
a = MinHash(20, 10, track_abundance=track_abundance)
with pytest.raises(TypeError):
a.merge(set())
