add a weekly ci schedule

.github/workflows/CI.yaml

@@ -7,20 +7,15 @@ on:
   pull_request:
     branches:
       - "main"
-  #schedule:
-  #  # Nightly tests run on master by default:
-  #  #   Scheduled workflows run on the latest commit on the default or base branch.
-  #  #   (from https://help.github.com/en/actions/reference/events-that-trigger-workflows#scheduled-events-schedule)
-  #  - cron: "0 0 * * *"
 
 jobs:
   test:
     name: Test on ${{ matrix.os }}, Python ${{ matrix.python-version }}
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        os: [ubuntu-latest] #[ubuntu-latest, macOS-latest]
-        python-version: ['3.10'] #[3.7, 3.8, 3.9, 3.10, 3.11, 3.12]
+        os: [ubuntu-latest]
+        python-version: ['3.10']
 
     steps:
     - uses: actions/checkout@v4
@@ -32,10 +27,8 @@ jobs:
         df -h
         ulimit -a
 
-
     - uses: mamba-org/setup-micromamba@v1
       with:
-        #python-version: ${{ matrix.python-version }}
         create-args: >-
           python=${{ matrix.python-version }}
         init-shell: bash
@@ -47,12 +40,10 @@ jobs:
         pip install . --no-deps
         conda list
 
-
     - name: Run tests
       shell: bash -l {0}
       run: |
-        pip install . --no-deps
-        pytest -v --color=yes --cov=serenityff --cov-report=xml --run-slow serenityff/charge/tests/ tests/
+        pytest -v --color=yes --cov=serenityff --cov-report=xml --run-slow --tier max serenityff/charge/tests/ tests/
 
   analyze:
     name: Analyze

.github/workflows/scheduled.yaml

@@ -0,0 +1,43 @@
+name: scheduled-tests
+
+on:
+  schedule:
+    - cron: "0 2 * * 6"
+
+jobs:
+  test:
+    name: Run ${{ matrix.env-name }} environment
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - env-name: "min"
+            file: "tree_only_env.yml"
+            mark: "min"
+          - env-name: "med"
+            file: "min_environment.yml"
+            mark: "med"
+          - env-name: "max"
+            file: "environment.yml"
+            mark: "max"
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Setup Conda
+        uses: conda-incubator/setup-miniconda@v3
+        with:
+          activate-environment: test-env
+          environment-file: ${{ matrix.file }}
+
+      - name: End-to-End Install
+        shell: bash -l {0}
+        run: |
+          conda install -y pytest pytest-cov
+          pip install git+https://github.com/rinikerlab/DASH-tree.git@${{ github.sha }}"
+
+      - name: Run Tests
+        shell: bash -l {0}
+        run: |
+          pytest -v --color=yes --cov=serenityff --cov-report=xml --run-slow --tier "${{ matrix.mark }}" serenityff/charge/tests/ tests/

serenityff/charge/tests/test_gnn_extraction.py

@@ -104,18 +104,21 @@ def graph(cwd) -> CustomData:
     )
 
 
+@pytest.mark.max
 def test_getter_setter(explainer) -> None:
     with pytest.raises(TypeError):
         explainer.gnn_explainer = "asdf"
     assert isinstance(explainer.gnn_explainer, GNNExplainer)
     return
 
 
+@pytest.mark.max
 def test_load(model, statedict) -> None:
     assert all(a == b for a, b in zip(model.state_dict(), statedict))
     return
 
 
+@pytest.mark.max
 def test_explain_atom(explainer, graph) -> None:
     explainer.gnn_explainer.explain_node(
         node_idx=0,
@@ -150,6 +153,7 @@ def test_explain_atom(explainer, graph) -> None:
     return
 
 
+@pytest.mark.max
 def test_extractor_properties(extractor, model, model_path, statedict_path, explainer) -> None:
     extractor._set_model(model)
     assert isinstance(extractor.model, ChargeCorrectedNodeWiseAttentiveFP)
@@ -166,6 +170,7 @@ def test_extractor_properties(extractor, model, model_path, statedict_path, expl
     assert isinstance(extractor.model, NodeWiseAttentiveFP)
 
 
+@pytest.mark.max
 def test_split_sdf(cwd, sdf_path) -> None:
     Extractor._split_sdf(
         sdf_file=sdf_path,
@@ -178,6 +183,7 @@ def test_split_sdf(cwd, sdf_path) -> None:
     return
 
 
+@pytest.mark.max
 def test_job_id(cwd) -> None:
     with open(f"{cwd}/id.txt", "w") as f:
         f.write("sdcep ab ein \n sdf <12345> saoeb <sd>")
@@ -187,11 +193,13 @@ def test_job_id(cwd) -> None:
     return
 
 
+@pytest.mark.max
 def test_mol_from_sdf(sdf_path):
     mol = mols_from_sdf(sdf_file=sdf_path)[0]
     assert mol.GetNumBonds() == 19
 
 
+@pytest.mark.max
 def test_graph_from_mol(mol, num_atoms, num_bonds, formal_charge, smiles) -> None:
     with pytest.raises(ValueError):
         get_graph_from_mol(mol=mol, index=0, sdf_property_name=None)
@@ -209,6 +217,7 @@ def test_graph_from_mol(mol, num_atoms, num_bonds, formal_charge, smiles) -> Non
     return
 
 
+@pytest.mark.max
 @pytest.mark.parametrize("sdf_prop", [(None), ("MBIScharge")])
 def test_graph_from_mol_no_y(mol, num_atoms, num_bonds, formal_charge, smiles, sdf_prop) -> None:
     graph = get_graph_from_mol(mol=mol, index=0, sdf_property_name=sdf_prop, no_y=True)
@@ -223,13 +232,15 @@ def test_graph_from_mol_no_y(mol, num_atoms, num_bonds, formal_charge, smiles, s
     return
 
 
+@pytest.mark.max
 def test_arg_parser(args, sdf_path, statedict_path) -> None:
     args = Extractor._parse_filenames(args)
     assert args.sdffile == sdf_path
     assert args.mlmodel == statedict_path
     return
 
 
+@pytest.mark.max
 def test_script_writing(cwd) -> None:
     Extractor._write_worker(directory=cwd)
     Extractor._write_cleaner(directory=cwd)
@@ -239,11 +250,13 @@ def test_script_writing(cwd) -> None:
     return
 
 
+@pytest.mark.max
 def test_explainer_initialization(extractor, model) -> None:
     extractor._initialize_expaliner(model=model, epochs=1)
     return
 
 
+@pytest.mark.max
 def test_command_to_shell_file(cwd) -> None:
     command_to_shell_file("echo Hello World", f"{cwd}/test.sh")
     os.path.isfile(f"{cwd}/test.sh")
@@ -253,6 +266,7 @@ def test_command_to_shell_file(cwd) -> None:
     os.remove(f"{cwd}/test.sh")
 
 
+@pytest.mark.max
 def test_csv_handling(cwd, sdf_path, extractor, model):
     extractor._initialize_expaliner(model=model, epochs=1)
     outfile = f"{cwd}/sdftest/combined.csv"
@@ -268,6 +282,7 @@ def test_csv_handling(cwd, sdf_path, extractor, model):
     rmtree(f"{cwd}/sdftest")
 
 
+@pytest.mark.max
 def test_run_extraction_local(extractor, statedict_path, cwd, sdf_path) -> None:
     extractor.run_extraction_local(
         sdf_file=sdf_path,

serenityff/charge/tests/test_gnn_training.py

@@ -85,6 +85,7 @@ def trainer(model, optimizer):
     return trainer
 
 
+@pytest.mark.max
 def test_init_and_forward_model(model, graph) -> None:
     model = model
     model.train()
@@ -99,6 +100,7 @@ def test_init_and_forward_model(model, graph) -> None:
     return
 
 
+@pytest.mark.max
 def test_initialize_trainer(trainer, model, sdf_path, pt_path, statedict_path, model_path, statedict) -> None:
     # test init
     assert trainer.device == device("cuda") if is_available() else device("cpu")
@@ -147,6 +149,7 @@ def test_initialize_trainer(trainer, model, sdf_path, pt_path, statedict_path, m
     return
 
 
+@pytest.mark.max
 def test_prepare_train_data(trainer, sdf_path):
     with pytest.warns(Warning):
         trainer.prepare_training_data()
@@ -158,6 +161,7 @@ def test_prepare_train_data(trainer, sdf_path):
     return
 
 
+@pytest.mark.max
 def test_train_model(trainer, sdf_path) -> None:
     trainer.gen_graphs_from_sdf(sdf_path)
     trainer.prepare_training_data(train_ratio=0.5)
@@ -173,6 +177,7 @@ def test_train_model(trainer, sdf_path) -> None:
     return
 
 
+@pytest.mark.max
 def test_prediction(trainer, graph, molecule) -> None:
 
     a = trainer.predict(graph)
@@ -198,6 +203,7 @@ def test_prediction(trainer, graph, molecule) -> None:
     return
 
 
+@pytest.mark.max
 def test_on_gpu(trainer) -> None:
 
     assert trainer._on_gpu == is_available()

serenityff/charge/tests/test_gnn_utils.py

@@ -126,6 +126,7 @@ def node_features() -> np.ndarray:
     )
 
 
+@pytest.mark.max
 def test_get_split_numbers() -> None:
     assert [1, 0] == get_split_numbers(N=1, train_ratio=0.5)
     assert [1, 1] == get_split_numbers(N=2, train_ratio=0.5)
@@ -136,13 +137,15 @@ def test_get_split_numbers() -> None:
     return
 
 
+@pytest.mark.max
 def test_random_split(data) -> None:
     train, test = split_data_random(data_list=data, train_ratio=0.5)
     assert len(train) == 6
     assert len(test) == 5
     return
 
 
+@pytest.mark.max
 def test_kfold_split(data) -> None:
     train1, test1 = split_data_Kfold(data, n_splits=2, split=0)
     train2, test2 = split_data_Kfold(data, n_splits=2, split=1)
@@ -151,6 +154,7 @@ def test_kfold_split(data) -> None:
     return
 
 
+@pytest.mark.max
 def test_initialization() -> None:
     featurizer = Featurizer()
     featurizer = MolecularFeaturizer()
@@ -159,13 +163,15 @@ def test_initialization() -> None:
     return
 
 
+@pytest.mark.max
 def test_one_hot_encode(atoms, allowable_set) -> None:
     assert one_hot_encode(atoms[0].GetSymbol(), allowable_set) == [1.0, 0.0, 0.0]
     assert one_hot_encode(atoms[0].GetSymbol(), allowable_set, include_unknown_set=True) == [1.0, 0.0, 0.0, 0.0]
     assert one_hot_encode(atoms[9].GetSymbol(), allowable_set, include_unknown_set=False) == [0.0, 0.0, 0.0]
     return
 
 
+@pytest.mark.max
 def test_hbond_constructor(mol) -> None:
     factory = _ChemicalFeaturesFactory.get_instance()
     import os
@@ -180,12 +186,14 @@ def test_hbond_constructor(mol) -> None:
     return
 
 
+@pytest.mark.max
 def test_H_bonding(mol, atoms) -> None:
     hbond_infos = construct_hydrogen_bonding_info(mol)
     assert get_atom_hydrogen_bonding_one_hot(atoms[11], hbond_infos) == [1.0, 1.0]
     return
 
 
+@pytest.mark.max
 def test_degree(atoms) -> None:
     assert np.where(get_atom_total_degree_one_hot(atoms[20])) == np.array([[1]])
     assert np.where(get_atom_total_degree_one_hot(atoms[11])) == np.array([[2]])
@@ -194,6 +202,7 @@ def test_degree(atoms) -> None:
     return
 
 
+@pytest.mark.max
 def test_atom_feature(mol, atoms, allowable_set) -> None:
     hbond_infos = construct_hydrogen_bonding_info(mol)
     feat = _construct_atom_feature(
@@ -206,6 +215,7 @@ def test_atom_feature(mol, atoms, allowable_set) -> None:
     return
 
 
+@pytest.mark.max
 def test_bond_feat(bonds) -> None:
     np.testing.assert_array_equal(np.where(_construct_bond_feature(bonds[0])), np.array([[3, 4, 5, 6]]))
     np.testing.assert_array_equal(np.where(_construct_bond_feature(bonds[6])), np.array([[0, 4, 6]]))
@@ -214,6 +224,7 @@ def test_bond_feat(bonds) -> None:
     return
 
 
+@pytest.mark.max
 def test_feature_vector_generation(smiles, mol, allowable_set, empty_set) -> None:
     featurizer = MolGraphConvFeaturizer(use_edges=True)
 
@@ -231,6 +242,7 @@ def test_feature_vector_generation(smiles, mol, allowable_set, empty_set) -> Non
     return
 
 
+@pytest.mark.max
 def test_graph_data_basics(node_features, edge_index, edge_features, node_pos_features) -> None:
     # Trigger all __init__() failures.
     with pytest.raises(TypeError):
@@ -255,6 +267,7 @@ def test_graph_data_basics(node_features, edge_index, edge_features, node_pos_fe
     return
 
 
+@pytest.mark.max
 def test_getters(
     graph_data,
     custom_graph_data,
@@ -285,6 +298,7 @@ def test_getters(
     return
 
 
+@pytest.mark.max
 def test_to_pyg(
     custom_graph_data,
     node_features,
@@ -301,6 +315,7 @@ def test_to_pyg(
     return
 
 
+@pytest.mark.max
 def test_custom_data_attributes(custom_data) -> None:
     data = custom_data
     assert data.smiles == "abc"
@@ -321,6 +336,7 @@ def test_custom_data_attributes(custom_data) -> None:
     return
 
 
+@pytest.mark.max
 def test_base_featurizer():
     featurizer = Featurizer()
     datapoints = [1]
@@ -332,6 +348,7 @@ def test_base_featurizer():
     assert np.array_equal(features, np.asarray([np.array([])]))
 
 
+@pytest.mark.max
 def test_molecular_featurizer(mol, smiles):
     featurizer = MolecularFeaturizer()
     featurizer.featurize([mol])
@@ -340,6 +357,7 @@ def test_molecular_featurizer(mol, smiles):
     featurizer.featurize(smiles)
 
 
+@pytest.mark.max
 def test_exceptions():
     with pytest.raises(NotInitializedError):
         raise NotInitializedError("msg")