def test_name(self):
# tests for Region.from_string are in docs
# here we test round-tripping
cases = [['contig1:50-100'] * 2, ['contig1:50-'] * 2,
['contig1:-100', 'contig1:0-100'], ['contig1', 'contig1:0-']]
for orig, parsed in cases:
a = Region.from_string(orig)
self.assertEqual(a.name, parsed)
def test_101_spam(self):
regions = [get_bam_regions(self.bam)[0]] * 100
regions += [Region.from_string("ref:0-100")] * 1000 # spam
self._run_one(19,
250,
0,
exp_batches=106,
exp_samples=2000,
exp_remains=1000,
regions=regions)
def test_100_convoluted(self):
# the chunk generator will give a larger overlap in the last chunk
# to give 4 chunks per region and no remainders for this:
regions = [get_bam_regions(self.bam)[0]] * 5
# but augment with 7 small subregions that will fall through
# as remainders:
regions += [Region.from_string("ref:0-1000")] * 7
self._run_one(2,
1300,
0,
exp_batches=10,
exp_samples=20,
exp_remains=7,
regions=regions)
def test_000_split_gap(self):
# The gapped bam has:
# @SQ SN:ref LN:30
# seq1 0 ref 1 7 10M
# seq2 0 ref 15 13 16M
# so an alignment from [0:10] and one from [14:30] without insertions
chunk_lengths = [10, 16]
region = Region.from_string('ref:0-30')
results = medaka.features.pileup_counts(region, __gapped_bam__)
self.assertEqual(len(results), 2, 'Number of chunks from gapped alignment')
for exp_len, chunk in zip(chunk_lengths, results):
for i in (0, 1):
# check both pileup and positions
self.assertEqual(exp_len, len(chunk[i]))
def test_001_cases(self, FastaFile):
class Args:
def __init__(self, **kwargs):
for k, v in kwargs.items():
setattr(self, k, v)
files = {1:"utg1190", 2:"scaffold_117"}
FastaFile.return_value = Args(references=['scaffold_117', 'utg1190'], lengths=[45079626, 16772114])
temp = tempfile.NamedTemporaryFile()
args = Args(draft="", threads=1, output=temp.name, fillgaps=False)
for fid, region in files.items():
fname = os.path.join(os.path.dirname(__file__), "data", "test_stitch_{}.hdf".format(fid))
args.inputs = fname
args.regions=[Region.from_string(region)]
try:
medaka.stitch.stitch(args)
except Exception as e:
self.fail("Stitching raise and Exception:\n {}".format(e))
def main():
# Entry point for testing/checking
logging.basicConfig(format='[%(asctime)s - %(name)s] %(message)s', datefmt='%H:%M:%S', level=logging.INFO)
np.set_printoptions(precision=4, linewidth=100)
parser = argparse.ArgumentParser('medaka', formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('bam', help='alignment file.')
parser.add_argument('region', help='alignment region to sample.')
parser.add_argument('--print', action='store_true', help='print counts.')
parser.add_argument('--dtypes', nargs='+', help='perform a multi-datatype tests.')
parser.add_argument('--norm', nargs='+', help='additional normalisation tests. (total, fwd_rev)')
args = parser.parse_args()
region = Region.from_string(args.region)
kwargs={
'log_min': None,
'max_hp_len': 1,
'is_compressed': False,
'consensus_as_ref': False,
'ref_mode': None,
'with_depth': False,
}
def _print(samples):
if args.print:
for p, f in zip(samples.positions, samples.features):
print('{}\t{}\t0\t{}\t{}'.format(p[0], p[1], '\t'.join('{:.3f}'.format(x) if x>0.0 else '-' for x in f), sum(f)))
dtype_options = [('',)]
if args.dtypes is not None:
dtype_options.append(args.dtypes)
norm_options = [None, ]
if args.norm is not None:
norm_options.extend(args.norm)
for dtypes in dtype_options:
kwargs['dtypes'] = dtypes
for norm in norm_options:
kwargs['normalise'] = norm
print("###########################################################")
print(kwargs)
encoder = FeatureEncoder(**kwargs)
# py-style
t0=now()
samples = encoder.bam_to_sample(args.bam, region, force_py=True)[0]
t1=now()
if not samples.is_empty:
print(samples.features.shape)
_print(samples)
else:
print("Samples is empty")
print("---------------------")
# C-style
t2=now()
samples = encoder.bam_to_sample(args.bam, region)[0]
t3=now()
if not samples.is_empty:
print(samples.features.shape)
_print(samples)
else:
print("Samples is empty")
print("pysam time:", t1 - t0)
print("hts time:", t3 - t2)