def _read_sequences(filename):
cat_call = factory.new("cat")
cat_call.add_multiple_values((filename,))
cat_call = cat_call.finalized_call
cat = None
try:
cat = procs.open_proc(cat_call,
bufsize=io.DEFAULT_BUFFER_SIZE,
stderr=procs.PIPE,
stdout=procs.PIPE)
qualities = _collect_qualities(cat.stdout, filename)
return sampling.reservoir_sampling(qualities, 100000)
except:
if cat:
cat.kill()
cat.wait()
cat = None
raise
finally:
rc_cat = cat.wait() if cat else 0
if rc_cat:
message = "Error running 'paleomix cat':\n" \
" Unicat return-code = %i\n\n%s" \
% (rc_cat, cat.stderr.read())
raise NodeError(message)
def _read_sequences(file_type, filename, stats):
cat_call = factory.new("cat")
cat_call.add_multiple_values((filename, ))
cat_call = cat_call.finalized_call
cat = None
try:
cat = procs.open_proc(cat_call,
bufsize=io.DEFAULT_BUFFER_SIZE,
stderr=procs.PIPE,
stdout=procs.PIPE)
qualities = _collect_qualities(cat.stdout, file_type, filename, stats)
return sampling.reservoir_sampling(qualities, 100000)
except StandardError as error:
if cat:
try:
cat.kill()
except OSError:
pass
cat.wait()
cat = None
raise error
finally:
rc_cat = cat.wait() if cat else 0
if rc_cat:
message = "Error running 'paleomix cat':\n" \
" Unicat return-code = %i\n\n%s" \
% (rc_cat, cat.stderr.read())
raise NodeError(message)
def __init__(self, handle, downsample, included_references):
self._records = collections.defaultdict(list)
references = handle.references
if len(references) != len(included_references):
raise ValueError("Length of 'included_references' must match the "
"number of references in BAM file.")
records = _filter_records(handle)
for record in reservoir_sampling(records, downsample):
key = references[record.tid]
self._records[key].append(record)
self.references = references
self._records = dict(self._records)
for value in self._records.values():
value.sort(key=lambda rec: rec.pos)
def __init__(self, handle, downsample, included_references):
self._records = collections.defaultdict(list)
references = handle.references
if len(references) != len(included_references):
raise ValueError("Length of 'included_references' must match the "
"number of references in BAM file.")
records = _filter_records(handle)
for record in reservoir_sampling(records, downsample):
key = references[record.tid]
self._records[key].append(record)
self.references = references
self._records = dict(self._records)
for value in self._records.values():
value.sort(key=lambda rec: rec.pos)
def test_reservoir_sampling__downsample_to_non_number_raises_type_error():
with pytest.raises(TypeError):
sampling.reservoir_sampling(list(range(5)), "Eh?")
def test_reservoir_sampling__downsample_to_float_raises_type_error():
with pytest.raises(TypeError):
sampling.reservoir_sampling(list(range(5)), 1.0)
def test_reservoir_sampling__downsample_to_negative_raises_value_error():
with pytest.raises(ValueError):
sampling.reservoir_sampling(list(range(5)), -1)
def test_reservoir_sampling__downsample_to_zero():
result = sampling.reservoir_sampling(list(range(5)), 0)
assert result == []
def test_reservoir_sampling__upsample_equals_input():
result = sampling.reservoir_sampling(list(range(5)), 10)
assert result == list(range(5))
def test_reservoir_sampling__select_second_item():
rng = Mock(randint=lambda _min, _max: 0)
values = [1, 2]
result = sampling.reservoir_sampling(values, 1, rng)
assert result == [2]
def test_reservoir_sampling__select_first_item():
rng = flexmock(randint=lambda _min, _max: 1)
values = [1, 2]
result = sampling.reservoir_sampling(values, 1, rng)
assert_equal(result, [1])
def test_reservoir_sampling__downsample_to_zero():
result = sampling.reservoir_sampling(range(5), 0)
assert_equal(result, [])
def test_reservoir_sampling__upsample_equals_input():
result = sampling.reservoir_sampling(range(5), 10)
assert_equal(result, range(5))
def test_reservoir_sampling__select_first_item():
rng = flexmock(randint=lambda _min, _max: 1)
values = [1, 2]
result = sampling.reservoir_sampling(values, 1, rng)
assert_equal(result, [1])
def test_reservoir_sampling__downsample_to_zero():
result = sampling.reservoir_sampling(range(5), 0)
assert_equal(result, [])
def test_reservoir_sampling__upsample_equals_input():
result = sampling.reservoir_sampling(range(5), 10)
assert_equal(result, range(5))