def test_kosman_pairwise_by_chunk(self):
a = numpy.array([[-1, -1], [0, 0], [0, 1],
[0, 0], [0, 0], [0, 1], [0, 1],
[0, 1], [0, 0], [0, 0], [0, 1]])
b = numpy.array([[1, 1], [-1, -1], [0, 0],
[0, 0], [1, 1], [0, 1], [1, 0],
[1, 0], [1, 0], [0, 1], [1, 1]])
c = numpy.full(shape=(11, 2), fill_value=1, dtype=numpy.int16)
d = numpy.full(shape=(11, 2), fill_value=1, dtype=numpy.int16)
gts = numpy.stack((a, b, c, d), axis=0)
gts = numpy.transpose(gts, axes=(1, 0, 2)).astype(numpy.int16)
variations = VariationsArrays()
variations['/calls/GT'] = gts
expected = [0.33333333, 0.75, 0.75, 0.45, 0.45, 0.]
distance = calc_pairwise_distance(variations, chunk_size=None)
assert numpy.allclose(distance, expected)
distance = calc_pairwise_distance(variations, chunk_size=2)
assert numpy.allclose(distance, expected)
# With all missing
a = numpy.full(shape=(10, 2), fill_value=-1, dtype=numpy.int16)
b = numpy.full(shape=(10, 2), fill_value=-1, dtype=numpy.int16)
gts = numpy.stack((a, b), axis=0)
gts = numpy.transpose(gts, axes=(1, 0, 2)).astype(numpy.int16)
variations = VariationsArrays()
variations['/calls/GT'] = gts
distance = calc_pairwise_distance(variations)
assert numpy.isnan(distance[0])
# With missing in some chunks only
variations['/calls/GT'][:5, 0, :] = 1
variations['/calls/GT'][:5, 1, :] = 0
assert calc_pairwise_distance(variations)[0] == 1
assert calc_pairwise_distance(variations, chunk_size=3)[0] == 1
def test_matching_pairwise_by_chunk(self):
a = numpy.array([[-1, -1], [0, 0], [0, 1], [0, 0], [0, 0], [0, 1],
[0, 1], [0, 1], [0, 0], [0, 0], [0, 1]])
b = numpy.array([[1, 1], [-1, -1], [0, 0], [0, 0], [1, 1], [0, 1],
[1, 0], [1, 0], [1, 0], [0, 1], [1, 1]])
c = numpy.full(shape=(11, 2), fill_value=1, dtype=numpy.int16)
d = numpy.full(shape=(11, 2), fill_value=1, dtype=numpy.int16)
gts = numpy.stack((a, b, c, d), axis=0)
gts = numpy.transpose(gts, axes=(1, 0, 2)).astype(numpy.int16)
variations = VariationsArrays()
variations['/calls/GT'] = gts
expected = [0.444444, 0, 0, 0.3, 0.3, 1]
distance = calc_pairwise_distance(variations,
chunk_size=None,
method='matching')
assert numpy.allclose(distance, expected)
distance = calc_pairwise_distance(variations,
chunk_size=2,
method='matching')
assert numpy.allclose(distance, expected)
# With all missing
a = numpy.full(shape=(10, 2), fill_value=-1, dtype=numpy.int16)
b = numpy.full(shape=(10, 2), fill_value=-1, dtype=numpy.int16)
gts = numpy.stack((a, b), axis=0)
gts = numpy.transpose(gts, axes=(1, 0, 2)).astype(numpy.int16)
variations = VariationsArrays()
variations['/calls/GT'] = gts
distance = calc_pairwise_distance(variations, method='matching')
assert numpy.isnan(distance[0])
def test_matching_pairwise_by_chunk(self):
a = numpy.array([[-1, -1], [0, 0], [0, 1],
[0, 0], [0, 0], [0, 1], [0, 1],
[0, 1], [0, 0], [0, 0], [0, 1]])
b = numpy.array([[1, 1], [-1, -1], [0, 0],
[0, 0], [1, 1], [0, 1], [1, 0],
[1, 0], [1, 0], [0, 1], [1, 1]])
c = numpy.full(shape=(11, 2), fill_value=1, dtype=numpy.int16)
d = numpy.full(shape=(11, 2), fill_value=1, dtype=numpy.int16)
gts = numpy.stack((a, b, c, d), axis=0)
gts = numpy.transpose(gts, axes=(1, 0, 2)).astype(numpy.int16)
variations = VariationsArrays()
variations['/calls/GT'] = gts
expected = [0.444444, 0, 0, 0.3, 0.3, 1]
distance = calc_pairwise_distance(variations, chunk_size=None,
method='matching')
assert numpy.allclose(distance, expected)
distance = calc_pairwise_distance(variations, chunk_size=2,
method='matching')
assert numpy.allclose(distance, expected)
# With all missing
a = numpy.full(shape=(10, 2), fill_value=-1, dtype=numpy.int16)
b = numpy.full(shape=(10, 2), fill_value=-1, dtype=numpy.int16)
gts = numpy.stack((a, b), axis=0)
gts = numpy.transpose(gts, axes=(1, 0, 2)).astype(numpy.int16)
variations = VariationsArrays()
variations['/calls/GT'] = gts
distance = calc_pairwise_distance(variations, method='matching')
assert numpy.isnan(distance[0])
def test_kosman_pairwise_by_chunk(self):
a = numpy.array([[-1, -1], [0, 0], [0, 1], [0, 0], [0, 0], [0, 1],
[0, 1], [0, 1], [0, 0], [0, 0], [0, 1]])
b = numpy.array([[1, 1], [-1, -1], [0, 0], [0, 0], [1, 1], [0, 1],
[1, 0], [1, 0], [1, 0], [0, 1], [1, 1]])
c = numpy.full(shape=(11, 2), fill_value=1, dtype=numpy.int16)
d = numpy.full(shape=(11, 2), fill_value=1, dtype=numpy.int16)
gts = numpy.stack((a, b, c, d), axis=0)
gts = numpy.transpose(gts, axes=(1, 0, 2)).astype(numpy.int16)
variations = VariationsArrays()
variations['/calls/GT'] = gts
expected = [0.33333333, 0.75, 0.75, 0.45, 0.45, 0.]
distance = calc_pairwise_distance(variations,
chunk_size=None,
min_num_snps=1)
assert numpy.allclose(distance, expected)
distance = calc_pairwise_distance(variations, chunk_size=2)
assert numpy.allclose(distance, expected)
distance = calc_pairwise_distance(variations,
chunk_size=None,
min_num_snps=11)
assert numpy.sum(numpy.isnan(distance)) == 5
# With all missing
a = numpy.full(shape=(10, 2), fill_value=-1, dtype=numpy.int16)
b = numpy.full(shape=(10, 2), fill_value=-1, dtype=numpy.int16)
gts = numpy.stack((a, b), axis=0)
gts = numpy.transpose(gts, axes=(1, 0, 2)).astype(numpy.int16)
variations = VariationsArrays()
variations['/calls/GT'] = gts
distance = calc_pairwise_distance(variations)
assert numpy.isnan(distance[0])
# With missing in some chunks only
variations['/calls/GT'][:5, 0, :] = 1
variations['/calls/GT'][:5, 1, :] = 0
assert calc_pairwise_distance(variations)[0] == 1
assert calc_pairwise_distance(variations, chunk_size=3)[0] == 1
