Bug fix: the subset Schema now checks for instances of int Or float

CAT/base.py

@@ -219,15 +219,17 @@ def prep_core(core_df: SettingsDataFrame) -> SettingsDataFrame:
         # Checks the if the dummy is a string (atomic symbol) or integer (atomic number)
         formula = core.get_formula()
 
-        # Returns the indices and Atoms of all dummy atom ligand placeholders in the core
+        # Returns the indices of all dummy atom ligand placeholders in the core
         if not core.properties.dummies:
             at_idx = np.array([i for i, atom in enumerate(core) if atom.atnum == dummy])
         else:
-            at_idx = np.array([i for i in core.properties.dummies]) - 1
+            dummies = core.properties.dummies
+            at_idx = np.fromiter(dummies, count=len(dummies), dtype=int)
+            at_idx -= 1
         if subset:
             at_idx = distribute_idx(core, at_idx, **subset)
 
-        # Convert atomic indices into atoms
+        # Convert atomic indices into Atoms
         at_idx += 1
         at_idx.sort()
         core.properties.dummies = dummies = [core[i] for i in at_idx]
@@ -241,7 +243,9 @@ def prep_core(core_df: SettingsDataFrame) -> SettingsDataFrame:
         # Delete all core dummy atoms
         for at in dummies:
             core.delete_atom(at)
-        idx_tuples.append((formula, ' '.join(at_idx.astype(str))))
+        idx_tuples.append(
+            (formula, ' '.join(at_idx.astype(str)))
+        )
 
     # Create and return a new dataframe
     idx = pd.MultiIndex.from_tuples(idx_tuples, names=['formula', 'anchor'])

CAT/data_handling/validation_schemas.py

@@ -253,13 +253,16 @@ def _get_crsjob() -> type:
 #: Schema for validating the ``['optional']['core']['subset']`` block.
 subset_schema: Schema = Schema({
     'p':
-        And(int, float, lambda n: 0.0 < n <= 1.0, Use(float)),
+        Or(
+            And(int, lambda n: 0 < n <= 1, Use(float)),
+            And(float, lambda n: 0.0 < n <= 1.0)
+        ),
 
     Optional_('mode', default='uniform'):
         And(str, lambda n: n.lower() in {'uniform', 'random', 'cluster'}, Use(str.lower)),
 
     Optional_('start', default=None):
-        And(None, int)
+        Or(None, int)
 })
 
 _db_names = ('core', 'ligand', 'qd')

tests/test_schemas.py

@@ -13,7 +13,7 @@
 from CAT.utils import get_template
 from CAT.data_handling.validation_schemas import (
     mol_schema, core_schema, ligand_schema, qd_schema, database_schema,
-    mongodb_schema, bde_schema, qd_opt_schema, crs_schema
+    mongodb_schema, bde_schema, qd_opt_schema, crs_schema, subset_schema
 )
 
 PATH = join('tests', 'test_files')
@@ -440,3 +440,52 @@ def test_bde_schema() -> None:
         assertion.eq(bde_schema.validate(bde_dict)[s], ref)
         bde_dict[s] = join(PATH, 'settings.yaml')
         assertion.eq(bde_schema.validate(bde_dict)[s], ref)
+
+
+def test_subset_schema() -> None:
+    """Test :data:`CAT.data_handling.validation_schemas.subset_schema`."""
+    subset_dict = {'p': 0.5}
+    ref = {
+        'p': 0.5,
+        'mode': 'uniform',
+        'start': None
+    }
+
+    assertion.eq(subset_schema.validate(subset_dict), ref)
+
+    subset_dict['p'] = 'bob'  # Exception: incorrect type
+    assertion.assert_(subset_schema.validate, subset_dict, exception=SchemaError)
+    subset_dict['p'] = -1  # Exception: incorrect value
+    assertion.assert_(subset_schema.validate, subset_dict, exception=SchemaError)
+    subset_dict['p'] = 0  # Exception: incorrect value
+    assertion.assert_(subset_schema.validate, subset_dict, exception=SchemaError)
+    subset_dict['p'] = 1.5  # Exception: incorrect value
+    assertion.assert_(subset_schema.validate, subset_dict, exception=SchemaError)
+    subset_dict['p'] = 0.33
+    assertion.eq(subset_schema.validate(subset_dict)['p'], 0.33)
+    subset_dict['p'] = 1
+    assertion.eq(subset_schema.validate(subset_dict)['p'], 1.0)
+
+    subset_dict['mode'] = 1  # Exception: incorrect type
+    assertion.assert_(subset_schema.validate, subset_dict, exception=SchemaError)
+    subset_dict['mode'] = 'bob'  # Exception: incorrect value
+    assertion.assert_(subset_schema.validate, subset_dict, exception=SchemaError)
+    subset_dict['mode'] = 'random'
+    assertion.eq(subset_schema.validate(subset_dict)['mode'], 'random')
+    subset_dict['mode'] = 'uniform'
+    assertion.eq(subset_schema.validate(subset_dict)['mode'], 'uniform')
+    subset_dict['mode'] = 'cluster'
+    assertion.eq(subset_schema.validate(subset_dict)['mode'], 'cluster')
+
+    subset_dict['start'] = 1.0  # Exception: incorrect type
+    assertion.assert_(subset_schema.validate, subset_dict, exception=SchemaError)
+    subset_dict['start'] = 'bob'  # Exception: incorrect type
+    assertion.assert_(subset_schema.validate, subset_dict, exception=SchemaError)
+    subset_dict['start'] = None
+    assertion.is_(subset_schema.validate(subset_dict)['start'], None)
+    subset_dict['start'] = 42
+    assertion.eq(subset_schema.validate(subset_dict)['start'], 42)
+    subset_dict['start'] = -5
+    assertion.eq(subset_schema.validate(subset_dict)['start'], -5)
+    subset_dict['start'] = 0
+    assertion.eq(subset_schema.validate(subset_dict)['start'], 0)