def test_windowed_tajima_d(self):
from allel import windowed_tajima_d
pos = np.array([1, 11, 21, 31, 41])
# example with calculable value
ac = AlleleCountsArray([[1, 3], [2, 2], [3, 1], [1, 3], [2, 2]])
expect = np.array([0.168] * 3)
actual, _, _ = windowed_tajima_d(pos, ac, size=25, step=10)
assert_array_almost_equal(expect, actual, decimal=3)
# too few sites
actual, _, _ = windowed_tajima_d(pos, ac, size=15, step=10)
assert 4 == len(actual)
assert np.all(np.isnan(actual))
# too few segregating sites
ac = AlleleCountsArray([[4, 0], [2, 2], [3, 1], [4, 0], [2, 2]])
actual, _, _ = windowed_tajima_d(pos, ac, size=25, step=10)
assert 3 == len(actual)
assert np.all(np.isnan(actual))
# allow people to override if they really want to
expect = np.array([0.592] * 3)
actual, _, _ = windowed_tajima_d(pos,
ac,
size=25,
step=10,
min_sites=2)
assert_array_almost_equal(expect, actual, decimal=3)
def test_moving_tajima_d(self):
from allel import moving_tajima_d
# example with calculable value
ac = AlleleCountsArray([[1, 3],
[2, 2],
[3, 1],
[1, 3],
[2, 2]])
expect = np.array([0.168] * 3)
actual = moving_tajima_d(ac, size=3, step=1)
assert_array_almost_equal(expect, actual, decimal=3)
# too few sites
actual = moving_tajima_d(ac, size=2, step=1)
assert 4 == len(actual)
assert np.all(np.isnan(actual))
# too few segregating sites
ac = AlleleCountsArray([[4, 0],
[2, 2],
[3, 1],
[4, 0],
[2, 2]])
actual = moving_tajima_d(ac, size=3, step=1)
assert 3 == len(actual)
assert np.all(np.isnan(actual))
# allow people to override if they really want to
expect = np.array([0.592] * 3)
actual = moving_tajima_d(ac, size=3, step=1, min_sites=2)
assert_array_almost_equal(expect, actual, decimal=3)
def test_slice_types(self):
ac = AlleleCountsArray(allele_counts_data, dtype='u1')
# total slice
s = ac[:]
assert isinstance(s, AlleleCountsArray)
# row slice
s = ac[1:]
assert isinstance(s, AlleleCountsArray)
# col slice
s = ac[:, 1:]
assert isinstance(s, np.ndarray)
# row index
s = ac[0]
assert isinstance(s, np.ndarray)
# col index
s = ac[:, 0]
assert isinstance(s, np.ndarray)
# item
s = ac[0, 0]
assert isinstance(s, np.uint8)
def test_tajima_d(self):
from allel import tajima_d
# example with calculable value
ac = AlleleCountsArray([[1, 3], [2, 2], [3, 1]])
expect = approx(0.168, 0.01)
actual = tajima_d(ac)
assert expect == actual
# too few sites
ac = AlleleCountsArray([[2, 2], [3, 1]])
assert np.nan is tajima_d(ac)
# too few segregating sites
ac = AlleleCountsArray([[4, 0], [2, 2], [3, 1]])
assert np.nan is tajima_d(ac)
# allow people to override if they really want to
assert approx(0.592, 0.01) == tajima_d(ac, min_sites=2)
def test_sequence_divergence(self):
from allel import sequence_divergence
pos = [2, 4, 8]
ac1 = AlleleCountsArray([[2, 0], [2, 0], [2, 0]])
ac2 = AlleleCountsArray([[0, 2], [0, 2], [0, 2]])
# all variants
e = 3 / 7
a = sequence_divergence(pos, ac1, ac2)
assert e == a
# start/stop
e = 2 / 6
a = sequence_divergence(pos, ac1, ac2, start=0, stop=5)
assert e == a
# start/stop, an provided
an1 = ac1.sum(axis=1)
an2 = ac2.sum(axis=1)
e = 2 / 6
a = sequence_divergence(pos,
ac1,
ac2,
start=0,
stop=5,
an1=an1,
an2=an2)
assert e == a
def test_constructor(self):
# missing data arg
with pytest.raises(TypeError):
# noinspection PyArgumentList
AlleleCountsArray()
# data has wrong dtype
data = 'foo bar'
with pytest.raises(TypeError):
AlleleCountsArray(data)
# data has wrong dtype
data = [4., 5., 3.7]
with pytest.raises(TypeError):
AlleleCountsArray(data)
# data has wrong dimensions
data = [1, 2, 3]
with pytest.raises(TypeError):
AlleleCountsArray(data)
# data has wrong dimensions
data = diploid_genotype_data
with pytest.raises(TypeError):
AlleleCountsArray(data)
# valid data (typed)
ac = AlleleCountsArray(allele_counts_data, dtype='u1')
aeq(allele_counts_data, ac)
assert np.uint8 == ac.dtype
def test_reduce_types(self):
ac = AlleleCountsArray(allele_counts_data, dtype='u1')
for m in 'sum', 'max', 'argmax':
x = getattr(ac, m)(axis=1)
assert isinstance(x, np.ndarray)
def setup_instance(self, data):
return AlleleCountsArray(data)