def test_adding_workflow_components(factory_type):
"""
Test building workflows from a verity of workflow components.
"""
factory = factory_type()
# element filter
efilter = workflow_components.ElementFilter()
factory.add_workflow_component(efilter)
assert len(factory.workflow) == 1
# conformer generator
conformer_gen = workflow_components.StandardConformerGenerator()
conformer_gen.max_conformers = 200
factory.add_workflow_component(conformer_gen)
assert len(factory.workflow) == 2
# add element filter again and make sure the component name has been incremented
factory.add_workflow_component(efilter)
assert len(factory.workflow) == 3
assert efilter.component_name in factory.workflow
# try to add a non component
with pytest.raises(InvalidWorkflowComponentError):
factory.add_workflow_component(3)
with pytest.raises(ValidationError):
factory.workflow = {"first component": 3}
factory.workflow = {"test_conformer": conformer_gen}
assert len(factory.workflow) == 1
def test_conformer_apply(toolkit):
"""
Test applying the standard conformer generator to a workflow.
"""
toolkit_name, toolkit_class = toolkit
if toolkit_class.is_available():
conf_gen = workflow_components.StandardConformerGenerator()
conf_gen.toolkit = toolkit_name
conf_gen.max_conformers = 1
conf_gen.clear_existing = True
mols = get_stereoisomers()
# remove duplicates from the set
molecule_container = get_container(mols)
result = conf_gen.apply(molecule_container.molecules, processors=1)
assert result.component_name == conf_gen.component_name
assert result.component_description == conf_gen.dict()
# make sure each molecule has a conformer that passed
for molecule in result.molecules:
assert molecule.n_conformers == 1, print(molecule.conformers)
for molecule in result.filtered:
assert molecule.n_conformers == 0
else:
pytest.skip(f"Toolkit {toolkit_name} not available.")
def test_exporting_settings_workflow(file_type, factory_type):
"""
Test exporting the settings and a workflow to the different file types.
"""
with temp_directory():
factory = factory_type()
changed_attrs = {
"maxiter": 400,
"priority": "super_high",
"compute_tag": "test tag"
}
for attr, value in changed_attrs.items():
setattr(factory, attr, value)
conformer_gen = workflow_components.StandardConformerGenerator()
conformer_gen.max_conformers = 100
factory.add_workflow_component(conformer_gen)
file_name = "test." + file_type
factory.export_settings(file_name=file_name)
with open(file_name) as f:
data = f.read()
assert conformer_gen.component_name in data
assert str(conformer_gen.max_conformers) in data
def test_standardconformer_generator_validators(toolkit):
"""
Test the standard conformer generator which calls the OFFTK.
"""
toolkit_name, toolkit_class = toolkit
if toolkit_class.is_available():
conf_gen = workflow_components.StandardConformerGenerator()
# make sure we are casting to ints
conf_gen.max_conformers = 1.02
assert conf_gen.max_conformers == 1
# test the toolkit validator
with pytest.raises(ValueError):
conf_gen.toolkit = "ambertools"
conf_gen.toolkit = toolkit_name
assert toolkit_name in conf_gen.provenance()
conf_gen.clear_existing = "no"
assert conf_gen.clear_existing is False
assert "OpenforcefieldToolkit" in conf_gen.provenance()
else:
pytest.skip(f"Toolkit {toolkit} not available.")
def test_get_component():
"""
Make sure the correct component is returned when requested.
"""
gen = get_component("standardconformergenerator")
assert gen == workflow_components.StandardConformerGenerator()
def test_register_component_replace():
"""
Test registering a component that is already registered with and without using replace.
"""
# get the standard conformer generator
gen = workflow_components.StandardConformerGenerator(max_conformers=1)
with pytest.raises(ComponentRegisterError):
register_component(component=gen, replace=False)
# now register using replace with a new default
register_component(component=gen, replace=True)
gen2 = get_component(gen.component_name)
assert gen2.max_conformers == gen.max_conformers
# now return th list back to normal
register_component(component=workflow_components.StandardConformerGenerator(), replace=True)
def test_adding_multipule_workflow_components(factory_type):
"""
Test adding a list of workflow components.
"""
factory = factory_type()
efilter = workflow_components.ElementFilter()
weight = workflow_components.MolecularWeightFilter()
conformer = workflow_components.StandardConformerGenerator()
components = [efilter, weight, conformer]
factory.add_workflow_component(components)
assert len(factory.workflow) == 3
for component in components:
assert component.component_name in factory.workflow
def test_get_wrokflow_component(factory_type):
"""
Test retrieving a workflow component.
"""
factory = factory_type()
efilter = workflow_components.ElementFilter()
weight = workflow_components.MolecularWeightFilter()
conformer = workflow_components.StandardConformerGenerator()
components = [efilter, weight, conformer]
factory.add_workflow_component(components)
for component in components:
assert factory.get_workflow_component(
component.component_name) == component
def test_create_dataset(factory_dataset_type):
"""
Test making a the correct corresponding dataset type from a given factory type.
"""
factory = factory_dataset_type[0]()
element_filter = workflow_components.ElementFilter()
element_filter.allowed_elements = [1, 6, 8, 7]
factory.add_workflow_component(element_filter)
conformer_generator = workflow_components.StandardConformerGenerator(
max_conformers=1)
factory.add_workflow_component(conformer_generator)
mols = Molecule.from_file(get_data("tautomers_small.smi"),
"smi",
allow_undefined_stereo=True)
# set some settings
changed_attrs = {
"compute_tag": "test tag",
"dataset_tags": ["openff", "test"],
"maxiter": 400
}
for attr, value in changed_attrs.items():
setattr(factory, attr, value)
dataset = factory.create_dataset(dataset_name="test name",
molecules=mols,
description="Force field test",
tagline="A test dataset")
# check the attributes were changed
for attr, value in changed_attrs.items():
assert getattr(dataset, attr) == value
assert dataset.dataset_name == "test name"
assert isinstance(dataset, factory_dataset_type[1]) is True
# make sure molecules we filtered and passed
assert dataset.dataset != {}
assert dataset.filtered != {}
assert element_filter.component_name in dataset.filtered_molecules
def test_remove_workflow_componet(factory_type):
"""
Test removing a workflow component through the API.
"""
factory = factory_type()
efilter = workflow_components.ElementFilter()
weight = workflow_components.MolecularWeightFilter()
conformer = workflow_components.StandardConformerGenerator()
components = [efilter, weight, conformer]
factory.add_workflow_component(components)
assert len(factory.workflow) == 3
for component in components:
factory.remove_workflow_component(component.component_name)
assert factory.workflow == {}
def test_export_workflow_only(file_type, factory_type):
"""
Test exporting the workflow only from the factory.
"""
with temp_directory():
factory = factory_type()
conformer_gen = workflow_components.StandardConformerGenerator()
conformer_gen.max_conformers = 100
factory.add_workflow_component(conformer_gen)
file_name = "workflow." + file_type
factory.export_workflow(file_name)
with open(file_name) as workflow:
data = workflow.read()
assert "method" not in data
assert "basis" not in data
assert "tag" not in data
def test_clear_workflow(factory_type):
"""
Test clearing out the workflow.
"""
factory = factory_type()
efilter = workflow_components.ElementFilter()
weight = workflow_components.MolecularWeightFilter()
conformer = workflow_components.StandardConformerGenerator()
components = [efilter, weight, conformer]
factory.add_workflow_component(components)
factory.clear_workflow()
assert factory.workflow == {}
factory.add_workflow_component(components)
factory.workflow = {}
assert factory.workflow == {}
#!/usr/bin/env python3
import sys
from openforcefield.topology import Molecule
from qcsubmit import \
workflow_components # load in a list of workflow_components
from qcsubmit.factories import OptimizationDatasetFactory
if __name__ == "__main__":
breakpoint()
qcsds = OptimizationDatasetFactory()
component = workflow_components.StandardConformerGenerator()
component.max_conformers = 100
component.toolkit = "rdkit"
qcsds.add_workflow_component(component)
filter = workflow_components.RMSDCutoffConformerFilter()
filter.rmsd_cutoff = 4.0
qcsds.add_workflow_component(filter)
smi_file = sys.argv[1]
mols = [Molecule.from_smiles(smi.split()[0], allow_undefined_stereochemistry=True) for smi in open(smi_file).readlines()]
dataset = qcsds.create_dataset(dataset_name='my_dataset', molecules=mols, description="my test dataset.")
dataset.export_dataset("dataset.json")
# now lets save the workflow to json
qcsds.export_settings('output.yaml')
