def test_runs():
dregion = (10, 20)
i = [1 if dregion[0] <= x <= dregion[1] else 0 for x in range(100)]
assert runs(i, lambda x: x > 0) == [dregion]
assert runs_gte(i, 1) == [dregion]
assert runs_gte(i, 1, 20) == []
assert runs_gte([], 1) == []
assert runs([], lambda x: True) == []
i = [0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1]
ir = [(4, 8), (14, 16)]
assert runs(i, lambda x: x > 0) == ir
assert runs_gte(i, 1) == ir
assert runs_gte_uint8(np.array(i, dtype=np.uint8), 1) == ir
i = [0] * 100
assert runs(i, lambda x: x > 0) == []
assert runs_gte(i, 1) == []
assert runs_gte_uint8(np.uint8(i), 1) == []
i = [1] * 100
ir = [(0, 99)]
assert runs(i, lambda x: x > 0) == ir
assert runs_gte(i, 1) == ir
assert runs_gte_uint8(np.array(i, dtype=np.uint8), 1) == ir
def test_runs():
dregion = (10,20)
i = [1 if dregion[0] <= x <= dregion[1] else 0 for x in range(100)]
assert runs(i, lambda x: x>0) == [dregion]
assert runs_gte(i, 1) == [dregion]
assert runs_gte(i, 1, 20) == []
assert runs_gte([], 1) == []
assert runs([], lambda x: True) == []
i = [0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1]
ir = [(4,8), (14, 16)]
assert runs(i, lambda x: x > 0) == ir
assert runs_gte(i, 1) == ir
assert runs_gte_uint8(np.array(i, dtype=np.uint8), 1) == ir
i = [0] * 100
assert runs(i, lambda x: x > 0) == []
assert runs_gte(i, 1) == []
assert runs_gte_uint8(np.uint8(i),1) == []
i = [1] * 100
ir = [(0,99)]
assert runs(i, lambda x: x>0) == ir
assert runs_gte(i, 1) == ir
assert runs_gte_uint8(np.array(i, dtype=np.uint8), 1) == ir
def make_intervals(ibdarray):
ibdarray = np.array(ibdarray, dtype=np.uint8)
ibdarray = ibdarray.copy()
# Get the intervals that are IBD=2 and remove them from the array
ibd2_tracts = [x for x in runs_gte_uint8(ibdarray, 2)]
for start, stop in ibd2_tracts:
ibdarray[start:(stop + 1)] -= 1
# Now get the remaining IBD=1 tracts and remove them from the array
ibd1_tracts = [x for x in runs_gte_uint8(ibdarray, 1)]
for start, stop in ibd1_tracts:
ibdarray[start:(stop + 1)] -= 1
return ibd1_tracts + ibd2_tracts
def make_intervals(ibdarray):
ibdarray = np.array(ibdarray, dtype=np.uint8)
ibdarray = ibdarray.copy()
# Get the intervals that are IBD=2 and remove them from the array
ibd2_tracts = [x for x in runs_gte_uint8(ibdarray, 2)]
for start, stop in ibd2_tracts:
ibdarray[start:(stop + 1)] -= 1
# Now get the remaining IBD=1 tracts and remove them from the array
ibd1_tracts = [x for x in runs_gte_uint8(ibdarray, 1)]
for start, stop in ibd1_tracts:
ibdarray[start:(stop + 1)] -= 1
return ibd1_tracts + ibd2_tracts
def _process_segments(identical,
min_seg=100,
min_val=1,
chromobj=None,
min_density=100,
size_unit='mb',
min_length=1,
maxmiss=0.25):
# IBD segments are long runs of identical genotypes
ibd = runs_gte_uint8(identical, min_val, minlength=min_seg)
if not ibd:
return ibd
# Genotype errors are things that happen. If theres a small gap between
# two IBD segments, we'll chalk that up to a genotyping error and join
# them together.
ibd = join_gaps(ibd, max_gap=2)
if chromobj:
ibd = filter_segments(chromobj,
ibd,
identical,
min_length=min_length,
size_unit=size_unit,
min_density=min_density,
maxmiss=maxmiss)
return ibd
def _process_segments(identical, min_seg=100, min_val=1, chromobj=None,
min_density=100, size_unit='mb',
min_length=1, maxmiss=0.25):
# IBD segments are long runs of identical genotypes
ibd = runs_gte_uint8(identical, min_val, minlength=min_seg)
if not ibd:
return ibd
# Genotype errors are things that happen. If theres a small gap between
# two IBD segments, we'll chalk that up to a genotyping error and join
# them together.
ibd = join_gaps(ibd, max_gap=2)
if chromobj:
ibd = filter_segments(chromobj, ibd, identical,
min_length=min_length,
size_unit=size_unit,
min_density=min_density,
maxmiss=maxmiss)
return ibd
