[ADD] option of chirality to npclassifierfp + [FIX] multitask splitter has no option of seed working

Author: jcapels

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = deepmol
-version = 1.1.17
+version = 1.1.18
 description = DeepMol: a python-based machine and deep learning framework for drug discovery
 keywords = machine-learning, deep-learning, cheminformatics, drug-discovery
 author = DeepMol Team

src/deepmol/__init__.py

@@ -1,2 +1,2 @@
 
-__version__ = '1.1.17'
+__version__ = '1.1.18'

src/deepmol/compound_featurization/np_classifier_fp.py

@@ -7,9 +7,10 @@
 
 class NPClassifierFP(MolecularFeaturizer):
 
-    def __init__(self, radius: int = 2, **kwargs) -> None:
+    def __init__(self, radius: int = 2, useChirality: bool = False, **kwargs) -> None:
         super().__init__(**kwargs)
         self.radius = radius
+        self.useChirality = useChirality
         self.feature_names = [f'npclassifier_{i}' for i in range(2048*(self.radius+1))]
 
     def _featurize(self, mol: Mol) -> np.ndarray:
@@ -31,7 +32,7 @@ def _featurize(self, mol: Mol) -> np.ndarray:
 
         mol_bi = {}
         for r in range(self.radius+1):
-            mol_fp = rdMolDescriptors.GetMorganFingerprintAsBitVect(mol, radius=r, bitInfo=mol_bi, nBits = 2048)
+            mol_fp = rdMolDescriptors.GetMorganFingerprintAsBitVect(mol, radius=r, bitInfo=mol_bi, nBits = 2048, useChirality=self.useChirality)
             mol_bi_QC = []
             for i in mol_fp.GetOnBits():
                 num_ = len(mol_bi[i])

src/deepmol/splitters/multitask_splitter.py

@@ -44,21 +44,21 @@ def split(self, dataset: Dataset, frac_train: float = 0.8, frac_valid: float = 0
 
         if seed is not None:
             np.random.seed(seed)
-
+        
         if frac_valid == 0:
             stratifier = IterativeStratification(n_splits=2, order=1,
-                                                 sample_distribution_per_fold=[frac_test, frac_train])
+                                                 sample_distribution_per_fold=[frac_test, frac_train], random_state=seed)
             train_indexes, test_indexes = next(stratifier.split(dataset.smiles, dataset.y))
 
             return train_indexes, [], test_indexes
         else:
             stratifier = IterativeStratification(n_splits=2, order=1, sample_distribution_per_fold=[frac_test,
-                                                                                                    1 - frac_test])
+                                                                                                    1 - frac_test], random_state=seed)
             train_indexes, test_indexes = next(stratifier.split(dataset.smiles, dataset.y))
 
             new_frac_train = frac_train / (1 - frac_test)
             stratifier = IterativeStratification(n_splits=2, order=1,
-                                                 sample_distribution_per_fold=[1 - new_frac_train, new_frac_train])
+                                                 sample_distribution_per_fold=[1 - new_frac_train, new_frac_train], random_state=seed)
 
             new_train_indexes, valid_indexes = next(stratifier.split(dataset.smiles[train_indexes],
                                                                      dataset.y[train_indexes]))

tests/unit_tests/featurizers/test_np_classifier_fp.py

@@ -10,3 +10,8 @@ def test_featurize(self):
         dataset_rows_number = len(self.mock_dataset.mols)
         d = NPClassifierFP().featurize(self.mock_dataset)
         self.assertEqual(dataset_rows_number, d._X.shape[0])
+
+    def test_feature_with_chirality(self):
+        dataset_rows_number = len(self.mock_dataset.mols)
+        d = NPClassifierFP(useChirality=True).featurize(self.mock_dataset)
+        self.assertEqual(dataset_rows_number, d._X.shape[0])