def verify_round_trip(self, ts, version):
msprime.dump_legacy(ts, self.temp_file, version=version)
with silence_stderr():
tsp = msprime.load_legacy(self.temp_file)
self.verify_tree_sequences_equal(ts, tsp)
tsp.dump(self.temp_file)
tsp = msprime.load(self.temp_file)
self.verify_tree_sequences_equal(ts, tsp)
def test_v2_no_samples(self):
ts = multi_locus_with_mutation_example()
msprime.dump_legacy(ts, self.temp_file, version=2)
root = h5py.File(self.temp_file, "r+")
del root['samples']
root.close()
tsp = msprime.load_legacy(self.temp_file)
self.verify_tree_sequences_equal(ts, tsp)
def verify_round_trip(self, ts, version):
msprime.dump_legacy(ts, self.temp_file, version=version)
tsp = msprime.load_legacy(self.temp_file)
simplify = version < 10
self.verify_tree_sequences_equal(ts, tsp, simplify=simplify)
tsp.dump(self.temp_file)
tsp = msprime.load(self.temp_file)
self.verify_tree_sequences_equal(ts, tsp, simplify=simplify)
def verify_malformed_json_v2(self, ts, group_name, attr, bad_json):
msprime.dump_legacy(ts, self.temp_file, 2)
# Write some bad JSON to the provenance string.
root = h5py.File(self.temp_file, "r+")
group = root[group_name]
group.attrs[attr] = bad_json
root.close()
tsp = msprime.load_legacy(self.temp_file)
self.verify_tree_sequences_equal(ts, tsp)
def run_upgrade(args):
try:
tree_sequence = msprime.load_legacy(args.source,
args.remove_duplicate_positions)
except msprime.DuplicatePositionsError:
exit(
"Error: Duplicate mutation positions in the source file detected.\n\n"
"This is not supported in the current file format. Running \"upgrade -d\" "
"will remove these duplicate positions. However, this will result in loss "
"of data from the original file!")
tree_sequence.dump(args.destination)
def test_duplicate_mutation_positions_single_value(self):
ts = multi_locus_with_mutation_example()
for version in [2, 3]:
msprime.dump_legacy(ts, self.temp_file, version=version)
root = h5py.File(self.temp_file, "r+")
root['mutations/position'][:] = 0
root.close()
self.assertRaises(msprime.DuplicatePositionsError,
msprime.load_legacy, self.temp_file)
tsp = msprime.load_legacy(self.temp_file,
remove_duplicate_positions=True)
self.assertEqual(tsp.num_sites, 1)
sites = list(tsp.sites())
self.assertEqual(sites[0].position, 0)
def test_duplicate_mutation_positions(self):
ts = multi_locus_with_mutation_example()
for version in [2, 3]:
msprime.dump_legacy(ts, self.temp_file, version=version)
root = h5py.File(self.temp_file, "r+")
position = np.array(root['mutations/position'])
position[0] = position[1]
root['mutations/position'][:] = position
root.close()
self.assertRaises(msprime.DuplicatePositionsError,
msprime.load_legacy, self.temp_file)
tsp = msprime.load_legacy(self.temp_file,
remove_duplicate_positions=True)
self.assertEqual(tsp.num_sites, position.shape[0] - 1)
position_after = list(s.position for s in tsp.sites())
self.assertEqual(list(position[1:]), position_after)
def verify_round_trip(self, ts):
with tempfile.NamedTemporaryFile(prefix="msp_ff_") as f:
tmp = sys.stderr
try:
with open("/dev/null", "w") as devnull:
sys.stderr = devnull
# We silence stderr here because h5py dumps out some
# spurious # error messages. See
# https://github.com/h5py/h5py/issues/390
msprime.dump_legacy(ts, f.name)
tsp = msprime.load_legacy(f.name)
finally:
sys.stderr = tmp
self.assertEqual(ts.get_num_records(), tsp.get_num_records())
self.assertEqual(ts.get_sample_size(), tsp.get_sample_size())
self.assertEqual(ts.get_sequence_length(), tsp.get_sequence_length())
self.assertEqual(ts.get_num_mutations(), tsp.get_num_mutations())
j = 0
for r1, r2 in zip(ts.records(), tsp.records()):
self.assertEqual(r1.left, r2.left)
self.assertEqual(r1.right, r2.right)
self.assertEqual(r1.node, r2.node)
self.assertEqual(r1.time, r2.time)
self.assertEqual(r1.children, r2.children)
self.assertEqual(r1.population, r2.population)
j += 1
self.assertEqual(j, ts.get_num_records())
self.assertEqual(ts.get_num_trees(), tsp.get_num_trees())
j = 0
for m1, m2 in zip(ts.mutations(), tsp.mutations()):
self.assertEqual(m1.position, m2.position)
self.assertEqual(m1.node, m2.node)
j += 1
self.assertEqual(ts.get_num_nodes(), tsp.get_num_nodes())
for u in range(ts.get_sample_size()):
self.assertEqual(ts.get_population(u), tsp.get_population(u))
provenance = tsp.get_provenance()
if ts.get_num_mutations() > 0:
self.assertEqual(len(provenance), 3)
else:
self.assertEqual(len(provenance), 2)
for p in provenance:
self.assertIsInstance(json.loads(p), dict)
def verify_round_trip(self, ts):
with tempfile.NamedTemporaryFile(prefix="msp_ff_") as f:
tmp = sys.stderr
try:
with open("/dev/null", "w") as devnull:
sys.stderr = devnull
# We silence stderr here because h5py dumps out some
# spurious # error messages. See
# https://github.com/h5py/h5py/issues/390
msprime.dump_legacy(ts, f.name)
tsp = msprime.load_legacy(f.name)
finally:
sys.stderr = tmp
self.assertEqual(ts.get_num_records(), tsp.get_num_records())
self.assertEqual(ts.get_sample_size(), tsp.get_sample_size())
self.assertEqual(ts.get_sequence_length(), tsp.get_sequence_length())
self.assertEqual(ts.get_num_mutations(), tsp.get_num_mutations())
j = 0
for r1, r2 in zip(ts.records(), tsp.records()):
self.assertEqual(r1.left, r2.left)
self.assertEqual(r1.right, r2.right)
self.assertEqual(r1.node, r2.node)
self.assertEqual(r1.time, r2.time)
self.assertEqual(r1.children, r2.children)
self.assertEqual(r1.population, r2.population)
j += 1
self.assertEqual(j, ts.get_num_records())
self.assertEqual(ts.get_num_trees(), tsp.get_num_trees())
j = 0
for m1, m2 in zip(ts.mutations(), tsp.mutations()):
self.assertEqual(m1.position, m2.position)
self.assertEqual(m1.node, m2.node)
j += 1
self.assertEqual(ts.get_num_nodes(), tsp.get_num_nodes())
for u in range(ts.get_sample_size()):
self.assertEqual(ts.get_population(u), tsp.get_population(u))
provenance = tsp.get_provenance()
if ts.get_num_mutations() > 0:
self.assertEqual(len(provenance), 3)
else:
self.assertEqual(len(provenance), 2)
for p in provenance:
self.assertIsInstance(json.loads(p), dict)
def run_upgrade(args):
tree_sequence = msprime.load_legacy(args.source)
tree_sequence.dump(args.destination)
def test_msprime_v_0_3_0(self):
ts = msprime.load_legacy(
"tests/data/hdf5-formats/msprime-0.3.0_v2.0.hdf5")
self.verify_tree_sequence(ts)
def run_upgrade(args):
tree_sequence = msprime.load_legacy(args.source)
tree_sequence.dump(args.destination)