def test_record_retrieval(self, data, base_kmer_size, num_colors_per_graph,
n_kmers):
# given
kmer_size = base_kmer_size * 2 + 1
num_colors = sum(num_colors_per_graph)
collection_builder = GraphCollection(
n_colors_per_graph=num_colors_per_graph, kmer_size=kmer_size)
expected_kmers = []
seen_kmers = set()
for _ in range(n_kmers):
kmer = data.draw(kmer_records(kmer_size, num_colors))
while kmer.kmer in seen_kmers:
kmer = data.draw(kmer_records(kmer_size, num_colors))
seen_kmers.add(kmer.kmer)
collection_builder.with_kmer_record(kmer)
expected_kmers.append(kmer)
collection = collection_builder.build()
# when
for expected_kmer in expected_kmers:
kmer = collection[expected_kmer.kmer]
# then
assert expected_kmer.kmer == kmer.kmer
assert expected_kmer.coverage == kmer.coverage
assert expected_kmer.edges == kmer.edges
def test_record_retrieval(self, data, kmer_size, num_colors, n_kmers):
# given
assume(n_kmers <= 4**(kmer_size - 1))
graph_builder = (builder.Graph().with_kmer_size(
kmer_size).with_num_colors(num_colors))
expected_kmers = []
seen_kmers = set()
for _ in range(n_kmers):
kmer = data.draw(kmer_records(kmer_size, num_colors))
while kmer.kmer in seen_kmers:
kmer = data.draw(kmer_records(kmer_size, num_colors))
seen_kmers.add(kmer.kmer)
graph_builder.with_kmer_record(kmer)
expected_kmers.append(kmer)
cg = self.RAClass(graph_builder.build())
# when
for expected_kmer in expected_kmers:
kmer = cg[expected_kmer.kmer]
# then
assert expected_kmer.kmer == kmer.kmer
assert np.all(expected_kmer.coverage == kmer.coverage)
for expected, actual in zip(expected_kmer.edges, kmer.edges):
assert expected == actual
def test_index(self, data, kmer_size, n_kmers):
# given
assume(kmer_size % 2 == 1)
num_colors = 1
graph_builder = (builder.Graph().with_kmer_size(
kmer_size).with_num_colors(num_colors))
expected_kmers = []
seen_kmers = set()
for _ in range(n_kmers):
kmer = data.draw(kmer_records(kmer_size, num_colors))
while kmer.kmer in seen_kmers:
kmer = data.draw(kmer_records(kmer_size, num_colors))
seen_kmers.add(kmer.kmer)
graph_builder.with_kmer_record(kmer)
expected_kmers.append(kmer)
expected_kmers = sorted(expected_kmers)
graph_stream = graph_builder.build()
header_stream = graph_builder.header.build()
header = Header.from_stream(header_stream)
# when
sequence = KmerUintSequence(graph_handle=graph_stream,
body_start=len(header_stream.getvalue()),
header=header,
n_records=len(expected_kmers))
# then
for idx, expected_kmer in enumerate(expected_kmers):
# then
assert idx == sequence.index_kmer_string(expected_kmer.kmer)
def test_parses_records(self, data, kmer_size, num_colors, n_kmers,
test_serializer):
# given
assume(n_kmers <= (4**kmer_size) / 4)
graph_builder = (builder.Graph().with_kmer_size(
kmer_size).with_num_colors(num_colors))
expected_kmers = []
seen = set()
for _ in range(n_kmers):
kmer = data.draw(kmer_records(kmer_size, num_colors))
while kmer.kmer in seen:
kmer = data.draw(kmer_records(kmer_size, num_colors))
seen.add(kmer.kmer)
graph_builder.with_kmer_record(kmer)
expected_kmers.append(kmer)
ra_parser = self.RAClass(graph_builder.build())
if test_serializer:
for real_kmer in ra_parser.values():
buffer = io.BytesIO()
real_kmer.dump(buffer)
assert real_kmer._kmer_data._data == buffer.getvalue()
sample_names = ra_parser.sample_names
buffer = io.BytesIO()
key_list = list(ra_parser.keys())
random.shuffle(key_list)
kmer_serializer \
.Kmers(keys=key_list,
val_callable=lambda k: ra_parser[k],
kmer_size=kmer_size,
num_colors=num_colors,
sample_names=sample_names) \
.dump(buffer)
buffer.seek(0)
ra_parser = self.RAClass(buffer)
# when
for expected_kmer in expected_kmers:
kmer = ra_parser[expected_kmer.kmer]
# then
assert expected_kmer.kmer == kmer.kmer
assert np.all(expected_kmer.coverage == kmer.coverage)
assert expected_kmer.edges == kmer.edges
def test_converts_to_raw(self, data, kmer_size, num_colors, from_bioseq):
# given
assume(kmer_size % 2 == 1)
kmer = Kmer.from_kmer_data(
KmerData(data.draw(kmer_records(kmer_size,
num_colors)).to_bytestring(),
kmer_size=kmer_size,
num_colors=num_colors))
converter = StringKmerConverter(kmer.kmer_size)
# when
if from_bioseq:
raw_kmer = converter.to_raw(Seq(kmer.kmer))
else:
raw_kmer = converter.to_raw(kmer.kmer)
# then
assert kmer.get_raw_kmer() == raw_kmer