def test_select_good_scoring_models_extract(self):
"""Test select_good with extract"""
with IMP.test.temporary_directory() as tmpdir:
mod_dir = os.path.join(tmpdir, 'modeling')
shutil.copytree(self.get_input_file_name('modeling'), mod_dir)
self.run_python_module(
select_good,
['-rd', mod_dir, '-rp', 'run',
'-sl', 'CrossLinkingMassSpectrometryRestraint_Distance_',
'-pl', 'ConnectivityRestraint_Rpb1',
'CrossLinkingMassSpectrometryRestraint_Data_Score_Chen',
'ExcludedVolumeSphere_None', 'Total_Score',
'-alt', '1.0', '-aut', '1.0', '-mlt', '0.0', '-mut', '12.0',
'-e'])
gsm_dir = os.path.join(mod_dir, 'good_scoring_models')
for score, num in (('A', 5), ('B', 4)):
score_file = os.path.join(gsm_dir, 'scores%s.txt' % score)
with open(score_file) as fh:
wc = len(fh.readlines())
self.assertEqual(wc, num)
os.unlink(score_file)
model_ids = os.path.join(gsm_dir, 'model_ids_scores.txt')
with open(model_ids) as fh:
wc = len(fh.readlines())
self.assertEqual(wc, 10)
os.unlink(model_ids)
model_ids = os.path.join(gsm_dir, 'model_sample_ids.txt')
with open(model_ids) as fh:
lines = fh.readlines()
self.assertEqual(len(lines), 9)
for line in lines:
num, sample = line.rstrip('\r\n').split()
rmf = os.path.join(gsm_dir, 'sample_%s' % sample,
'%s.rmf3' % num)
if hasattr(RMF.NodeHandle, 'replace_child'):
r = RMF.open_rmf_file_read_only(rmf)
cpf = RMF.CombineProvenanceConstFactory(r)
fpf = RMF.FilterProvenanceConstFactory(r)
rn = r.get_root_node().get_children()[0]
# Should be one Provenance node
prov, = [n for n in rn.get_children()
if n.get_type() == RMF.PROVENANCE]
# Top-level provenance should be FilterProvenance
self.assertTrue(fpf.get_is(prov))
fp = fpf.get(prov)
self.assertEqual(fp.get_method(), "Best scoring")
self.assertEqual(fp.get_frames(), 9)
# Next provenance should be CombineProvenance
prov, = prov.get_children()
self.assertTrue(cpf.get_is(prov))
cp = cpf.get(prov)
self.assertEqual(cp.get_runs(), 2)
self.assertEqual(cp.get_frames(), 100)
os.unlink(rmf)
os.unlink(model_ids)
os.rmdir(os.path.join(gsm_dir, 'sample_A'))
os.rmdir(os.path.join(gsm_dir, 'sample_B'))
os.rmdir(gsm_dir)
def test_exhaust(self):
"""Test the master sampling exhaustiveness script"""
try:
import pyRMSD
except ImportError:
self.skipTest("this test requires the pyRMSD Python module")
with IMP.test.temporary_working_directory() as tmpdir:
self.make_models(tmpdir)
gsm_dir = os.path.join(tmpdir, 'modeling', 'good_scoring_models')
self.run_python_module(exhaust, [
'-n', 'test', '-p', gsm_dir, '-d',
self.get_input_file_name('density_ranges.txt'), '-m',
'cpu_omp', '-c', '8', '-a', '-g', '0.5', '-gp'
])
if hasattr(RMF.NodeHandle, 'replace_child'):
r = RMF.open_rmf_file_read_only(
os.path.join(tmpdir, 'cluster.0',
'cluster_center_model.rmf3'))
clpf = RMF.ClusterProvenanceConstFactory(r)
cpf = RMF.CombineProvenanceConstFactory(r)
fpf = RMF.FilterProvenanceConstFactory(r)
rn = r.get_root_node().get_children()[0]
# Should be one Provenance node
prov, = [
n for n in rn.get_children()
if n.get_type() == RMF.PROVENANCE
]
# Top-level provenance should be ClusterProvenance
self.assertTrue(clpf.get_is(prov))
cp = clpf.get(prov)
self.assertEqual(cp.get_members(), 36)
self.assertAlmostEqual(cp.get_precision(), 0.42, delta=0.01)
self.assertEqual(cp.get_density(),
os.path.abspath('cluster.0/LPD_TestAll.mrc'))
# Next provenance should be filter, combine
prov, = prov.get_children()
self.assertTrue(fpf.get_is(prov))
prov, = prov.get_children()
self.assertTrue(cpf.get_is(prov))
# Check for expected files
expected = [
'Distances_Matrix.data.npy', 'Identities_A.txt',
'Identities_B.txt', 'cluster.0.all.txt',
'cluster.0.sample_A.txt', 'cluster.0.sample_B.txt',
'test.ChiSquare.pdf', 'test.ChiSquare_Grid_Stats.txt',
'test.Cluster_Population.pdf', 'test.Cluster_Population.txt',
'test.Cluster_Precision.txt', 'test.KS_Test.txt',
'test.Sampling_Precision_Stats.txt', 'test.Score_Dist.pdf',
'test.Score_Hist_A.txt', 'test.Score_Hist_B.txt',
'test.Top_Score_Conv.pdf', 'test.Top_Score_Conv.txt',
'cluster.0/cluster_center_model.rmf3',
'cluster.0/LPD_TestAll.mrc',
'cluster.0/Sample_A/LPD_TestAll.mrc',
'cluster.0/Sample_B/LPD_TestAll.mrc'
]
for e in expected:
os.unlink(os.path.join(tmpdir, e))
os.rmdir(os.path.join(tmpdir, 'cluster.0', 'Sample_A'))
os.rmdir(os.path.join(tmpdir, 'cluster.0', 'Sample_B'))
os.rmdir(os.path.join(tmpdir, 'cluster.0'))
