def test_structureset_from_npz():
npz_path = './tests/data/md/6h2o.temelso.etal.pr.md.gfn2.300k.step10000-ase.md-orca.mp2.def2tzvp.300k.npz'
# Getting entity_ids.
h2o_size = 3
cluster_size = 6
entity_ids = []
for i in range(0, cluster_size):
entity_ids.extend([i for _ in range(0, h2o_size)])
# Getting comp_ids
solvent = 'h2o'
comp_ids = []
for i in range(0, cluster_size):
comp_ids.append(solvent)
# Creating structure set.
test_structureset = structureSet()
test_structureset.from_npz(npz_path, 'z', 'R', 'Angstrom', entity_ids,
comp_ids)
# Naming of the structure set.
assert test_structureset.name == '6h2o.temelso.etal.pr.md.gfn2.300k.step10000-ase.md-orca.mp2.def2tzvp.300k'
test_structureset.name = '6h2o.temelso.etal.pr.md.gfn2.300k.step10000-ase.md-orca.mp2.def2tzvp.300k'
example_6h2o_md(test_structureset)
def test_structureset_from_traj():
traj_path = './tests/data/md/10h2o.abc0.iter1.gfn2-xtb.md-gfn2.300k-1.traj'
# Getting entity_ids.
h2o_size = 3
cluster_size = 10
entity_ids = []
for i in range(0, cluster_size):
entity_ids.extend([i for _ in range(0, h2o_size)])
# Getting comp_ids
solvent = 'h2o'
comp_ids = []
for i in range(0, cluster_size):
comp_ids.append(solvent)
# Creating structure set.
test_structureset = structureSet()
test_structureset.from_xyz(traj_path, 'Angstrom', entity_ids, comp_ids)
# Naming of the structure set.
assert test_structureset.name == '10h2o.abc0.iter1.gfn2-xtb.md-gfn2.300k-1'
test_structureset.name = '10h2o.abc0.iter1.gfn2.md.gfn2.300k.iter1-mbgdml.structset'
# test_structureset.save(test_structureset.name, test_structureset.structureset, './tests/data/structuresets/')
example_10h2o(test_structureset)
def trim_140h2o_rset():
"""Trims the 140h2o structure set to make tests easier.
"""
n_R = 3 # Number of structures to keep.
n_entities = 5 # Number of molecules to keep in each structure.
molecule_size = molecule_sizes['h2o'] # Number of atoms in a water molecule.
rset = data.structureSet(rset_path_140h2o)
assert rset.type == 's'
assert rset.md5 == '3e5cf945f6b4e979373e7772410e3746'
# Trims and checks z.
rset.z = rset.z[:n_entities*molecule_size]
z = np.array([8, 1, 1, 8, 1, 1, 8, 1, 1, 8, 1, 1, 8, 1, 1])
assert np.all(rset.z == z)
# Trims and checks R.
rset.R = rset.R[:n_R, :molecule_size*n_entities]
r_2 = np.array([
[ 6.07124359, 0.7619846, 0.58984577],
[ 6.47807882, -0.18138608, 0.67938893],
[ 5.14951519, 0.76914325, 0.66198299],
[-4.28204826, -3.57395445, 0.81850038],
[-4.33819343, -4.29134079, 0.12722189],
[-4.33829705, -2.80167393, 0.40818626],
[-2.82371582, -3.52131402, -4.12086561],
[-2.96180787, -4.46433929, -3.79287547],
[-1.85909245, -3.46817877, -4.3649756,],
[ 6.24586283, -1.76605224, 0.72883595],
[ 5.51074538, -2.26847206, 1.21432844],
[ 6.92768826, -2.3359825, 0.25592583],
[-2.44826194, -6.14429515, -3.37660252],
[-2.19536627, -6.12210888, -2.51171765],
[-2.65953004, -7.04099688, -3.59504014]
])
assert np.allclose(rset.R[2], r_2)
assert rset.z.shape[0] == rset.R.shape[1]
# Trims and checks entity_ids.
rset.entity_ids = rset.entity_ids[:n_entities*molecule_size]
entity_ids = np.array([0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4])
assert np.all(rset.entity_ids == entity_ids)
# Trims and checks comp_ids
rset.comp_ids = rset.comp_ids[:n_entities]
comp_ids = np.array(['h2o', 'h2o', 'h2o', 'h2o', 'h2o'])
assert np.all(rset.comp_ids == comp_ids)
# Confirms changes with MD5.
assert rset.md5 == 'e6a7a058b5fefb622fb3296e29a84150'
return rset
def test_sample_dset_same_size():
"""
"""
dset_h2o_2body_path = f'{dset_dir}/2h2o/140h2o.sphere.gfn2.md.500k.prod1.3h2o.dset.2h2o-dset.mb.npz'
dset_h2o_2body = data.dataSet(dset_h2o_2body_path)
# Trim dset_h2o_2body to 50 structures
remaining = 50
for key in ['r_prov_specs', 'E', 'R', 'F']:
setattr(dset_h2o_2body, key, getattr(dset_h2o_2body, key)[:remaining])
dset_h2o_2body_cm_6 = data.dataSet()
dset_h2o_2body_cm_6.name = '140h2o.sphere.gfn2.md.500k.prod1.3h2o.dset.2h2o-dset.mb-cm.6'
dset_h2o_2body_cm_6 = dset_sample_structures(
dset_h2o_2body_cm_6, dset_h2o_2body, 'all', 2, criteria.cm_distance_sum,
np.array([]), np.array([6.0]), True, False
)
assert dset_h2o_2body_cm_6.theory == 'mp2.def2tzvp.frozencore'
assert dset_h2o_2body_cm_6.criteria == 'cm_distance_sum'
assert np.array_equal(dset_h2o_2body_cm_6.z_slice, np.array([]))
assert np.array_equal(dset_h2o_2body_cm_6.cutoff, np.array([6.0]))
assert np.array_equal(dset_h2o_2body_cm_6.entity_ids, np.array([0, 0, 0, 1, 1, 1]))
assert np.array_equal(
dset_h2o_2body_cm_6.comp_ids, np.array(['h2o', 'h2o'])
)
assert dset_h2o_2body_cm_6.centered == True
assert dset_h2o_2body_cm_6.r_unit == 'Angstrom'
# 8726c482c19cdf7889cd1e62b9e9c8e1 is the MD5 has for the full 140h2o rset.
assert dset_h2o_2body_cm_6.r_prov_ids == {0: '8726c482c19cdf7889cd1e62b9e9c8e1'}
assert np.array_equal(dset_h2o_2body_cm_6.z, np.array([8, 1, 1, 8, 1, 1]))
rset = data.structureSet(rset_path_140h2o)
check_R_with_rset(dset_h2o_2body_cm_6, rset, True)
# Checking energies and forces.
dset_r_prov_specs = dset_h2o_2body_cm_6.r_prov_specs
dset_E = dset_h2o_2body_cm_6.E
dset_F = dset_h2o_2body_cm_6.F
dset_sample_r_prov_specs = dset_h2o_2body.r_prov_specs
dset_sample_E = dset_h2o_2body.E
dset_sample_F = dset_h2o_2body.F
for i_r in range(len(dset_h2o_2body_cm_6.R)):
i_r_dset_sample = np.where(
np.all(dset_sample_r_prov_specs == dset_r_prov_specs[i_r], axis=1)
)[0][0]
assert np.allclose(dset_E[i_r], dset_sample_E[i_r_dset_sample])
assert np.allclose(dset_F[i_r], dset_sample_F[i_r_dset_sample])
def load_140h2o_rset():
return structureSet(Rset_140h2o_path)
def test_structureset_load():
structureset_path = './tests/data/structuresets/10h2o.abc0.iter1.gfn2.md.gfn2.300k.iter1-mbgdml.structset.npz'
test_structureset = structureSet(structureset_path)
example_10h2o(test_structureset)