feat: Integration of torch models in main

.gitignore

@@ -139,3 +139,9 @@ jaqpotpy/models/.DS_Store
 jaqpotpy/models/.DS_Store
 .DS_Store
 jaqpotpy/models/tests/.DS_Store
+
+*.pkl
+*.json
+venv-jaqpotpy/
+*.txt
+*.pt

docs_jaqpotpy_torch/jaqpotpy_torch/datasets/index.md

@@ -0,0 +1,8 @@
+Module jaqpotpy_torch.datasets
+==============================
+Author: Ioannis Pitoskas (jpitoskas@gmail.com)
+
+Sub-modules
+-----------
+* jaqpotpy_torch.datasets.smiles_graph_dataset
+* jaqpotpy_torch.datasets.tabular_dataset

docs_jaqpotpy_torch/jaqpotpy_torch/datasets/smiles_graph_dataset.md

@@ -0,0 +1,111 @@
+Module jaqpotpy_torch.datasets.smiles_graph_dataset
+===================================================
+Author: Ioannis Pitoskas (jpitoskas@gmail.com)
+
+Classes
+-------
+
+`SmilesGraphDataset(smiles, y=None, featurizer=None)`
+:   A PyTorch Dataset class for handling SMILES strings as graphs.
+    This class overrides `__getitem__` and `__len__` (check source code for methods' docstrings).
+
+    Attributes:
+        smiles (list): A list of SMILES strings.
+        y (list, optional): A list of target values.
+        featurizer (SmilesGraphFeaturizer): The object to transform SMILES strings into graph representations.
+        precomputed_features (list, optional): A list of precomputed features. If precompute_featurization() is not called, this attribute remains None.
+
+    The SmilesGraphDataset constructor.
+
+    Args:
+        smiles (list): A list of SMILES strings. 
+        y (list, optional): A list of target values. Default is None.
+        featurizer (SmilesGraphFeaturizer, optional): A featurizer object for to create graph representations from SMILES strings.
+
+    Example:
+    ```
+    >>> from jaqpotpy.jaqpotpy_torch.featurizers import SmilesGraphFeaturizer
+    >>> from rdkit import Chem
+    >>>
+    >>> smiles = ['C1=CN=CN1', 'CCCCCCC']
+    >>> y = [0, 1]
+    >>> featurizer = SmilesGraphFeaturizer()
+    >>> featurizer.add_atom_characteristic('symbol', ['C', 'O', 'Na', 'Cl', 'UNK'])
+    >>> featurizer.add_atom_characteristic('is_in_ring')
+    >>> featurizer.add_bond_characteristic('bond_type', [Chem.rdchem.BondType.SINGLE, Chem.rdchem.BondType.DOUBLE, Chem.rdchem.BondType.TRIPLE])
+    >>>
+    >>> dataset = SmilesGraphDataset(smiles, y=y, featurizer=featurizer)
+    >>> dataset[0]
+    Data(x=[5, 6], edge_index=[2, 10], edge_attr=[10, 3], y=0, smiles='C1=CN=CN1')
+    ```
+
+    ### Ancestors (in MRO)
+
+    * torch.utils.data.dataset.Dataset
+    * typing.Generic
+
+    ### Descendants
+
+    * jaqpotpy_torch.datasets.smiles_graph_dataset.SmilesGraphDatasetWithExternal
+
+    ### Methods
+
+    `get_atom_feature_labels(self)`
+    :   Returns the atom feature labels.
+
+        Returns:
+            list: A list of atom feature labels.
+
+    `get_bond_feature_labels(self)`
+    :   Returns the bond feature labels.
+
+        Returns:
+            list: A list of bond feature labels.
+
+    `get_num_edge_features(self)`
+    :   Returns the number of edge features.
+
+        Returns:
+            int: Number of edge features.
+
+    `get_num_node_features(self)`
+    :   Returns the number of node features.
+
+        Returns:
+            int: Number of node features.
+
+    `precompute_featurization(self)`
+    :   Precomputes the featurization of the dataset.
+
+`SmilesGraphDatasetWithExternal(smiles, external, y=None, featurizer=None)`
+:   A PyTorch Dataset class for handling SMILES strings as graphs, along with additional external features.
+    This class inherits from SmilesGraphDataset and overrides `__getitem__` and `__len__` (check source code for methods' docstrings).
+
+    Attributes:
+        smiles (list): A list of SMILES strings.
+        y (list, optional): A list of target values.
+        featurizer (SmilesGraphFeaturizer): The object to transform SMILES strings into graph representations.
+        precomputed_features (list, optional): A list of precomputed features. If precompute_featurization() is not called, this attribute remains None.
+        external (torch.Tensor): A 2D tensor containing the external features.
+
+    The SmilesGraphDatasetWithExternal constructor.
+
+    Args:
+        smiles (list): A list of SMILES strings.
+        external (numpy.ndarray or pandas.DataFrame): External feature data 2D matrix.
+        y (list, optional): A list of target values. Default is None.
+        featurizer (SmilesGraphFeaturizer, optional): A featurizer object for to create graph representations from SMILES strings. Default is None.
+
+    ### Ancestors (in MRO)
+
+    * jaqpotpy_torch.datasets.smiles_graph_dataset.SmilesGraphDataset
+    * torch.utils.data.dataset.Dataset
+    * typing.Generic
+
+    ### Methods
+
+    `get_num_external_features(self)`
+    :   Returns the number of external features.
+
+        Returns:
+            int: Number of external features.

docs_jaqpotpy_torch/jaqpotpy_torch/datasets/tabular_dataset.md

@@ -0,0 +1,42 @@
+Module jaqpotpy_torch.datasets.tabular_dataset
+==============================================
+Author: Ioannis Pitoskas (jpitoskas@gmail.com)
+
+Classes
+-------
+
+`TabularDataset(X, y=None)`
+:   A PyTorch Dataset class for handling tabular data.
+
+    Attributes:
+        X (torch.tensor): A 2D tensor containing the feature data.
+        y (torch.tensor): A 1D tensor containing the target data.
+
+    The TabularDataset constructor.
+
+    Args:
+        X (numpy.ndarray or pandas.DataFrame): Feature data matrix of shape (n_samples, n_features).
+        y (numpy.ndarray or pandas.DataFrame, optional): Target data of shape (n_samples,).
+
+    Example:
+    ```
+    >>> import numpy as np
+    >>> X = np.random.rand(3, 2)
+    >>> y = np.random.rand(3, 2)
+    >>> dataset = TabularDataset(X, y=y)
+    >>> dataset[0]
+    (tensor([0.7778, 0.3400]), tensor(0.4730))
+    ```
+
+    ### Ancestors (in MRO)
+
+    * torch.utils.data.dataset.Dataset
+    * typing.Generic
+
+    ### Methods
+
+    `get_num_features(self)`
+    :   Returns the number of features in the dataset.
+
+        Returns:
+            int: Number of features.

docs_jaqpotpy_torch/jaqpotpy_torch/featurizers/featurizer.md

@@ -0,0 +1,25 @@
+Module jaqpotpy_torch.featurizers.featurizer
+============================================
+Author: Ioannis Pitoskas (jpitoskas@gmail.com)
+
+Classes
+-------
+
+`Featurizer()`
+:   Abstract base class for featurizers.
+
+    ### Ancestors (in MRO)
+
+    * abc.ABC
+
+    ### Descendants
+
+    * jaqpotpy_torch.featurizers.smiles_graph_featurizer.SmilesGraphFeaturizer
+
+    ### Methods
+
+    `featurize(self, *args, **kwargs)`
+    :   Abstract method to featurize the input data.
+
+        Returns:
+            The featurized data.

docs_jaqpotpy_torch/jaqpotpy_torch/featurizers/index.md

@@ -0,0 +1,8 @@
+Module jaqpotpy_torch.featurizers
+=================================
+Author: Ioannis Pitoskas (jpitoskas@gmail.com)
+
+Sub-modules
+-----------
+* jaqpotpy_torch.featurizers.featurizer
+* jaqpotpy_torch.featurizers.smiles_graph_featurizer