0.24.20

manydataset

Author: bartzbeielstein

pyproject.toml

@@ -7,7 +7,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "spotpython"
-version = "0.24.19"
+version = "0.24.20"
 authors = [
   { name="T. Bartz-Beielstein", email="tbb@bartzundbartz.de" }
 ]

src/spotpython/data/manydataset.py

@@ -0,0 +1,106 @@
+from torch.utils.data import Dataset
+import torch
+import pandas as pd
+from typing import List, Optional, Union
+
+
+class ManyToManyDataset(Dataset):
+    """
+    A PyTorch Dataset for many-to-many data.
+
+    Args:
+        df_list (List[pd.DataFrame]): List of pandas DataFrames.
+        target (str): The target column name.
+        drop (Optional[Union[str, List[str]]]): Column(s) to drop from the DataFrames. Default is None.
+        dtype (torch.dtype): Data type for the tensors. Default is torch.float32.
+
+    Attributes:
+        data (List[pd.DataFrame]): List of pandas DataFrames with specified columns dropped.
+        target (List[torch.Tensor]): List of target tensors.
+        features (List[torch.Tensor]): List of feature tensors.
+
+    Examples:
+        >>> import pandas as pd
+        >>> from spotpython.data.manydataset import ManyToManyDataset
+        >>> df1 = pd.DataFrame({'feature1': [1, 2], 'feature2': [3, 4], 'target': [5, 6]})
+        >>> df2 = pd.DataFrame({'feature1': [7, 8], 'feature2': [9, 10], 'target': [11, 12]})
+        >>> dataset = ManyToManyDataset([df1, df2], target='target', drop='feature2')
+        >>> len(dataset)
+        2
+        >>> dataset[0]
+        (tensor([[1.],
+                 [2.]]), tensor([5., 6.]))
+    """
+
+    def __init__(
+        self,
+        df_list: List[pd.DataFrame],
+        target: str,
+        drop: Optional[Union[str, List[str]]] = None,
+        dtype: torch.dtype = torch.float32,
+    ):
+        try:
+            self.data = [df.drop(drop, axis=1) for df in df_list]
+        except KeyError:
+            self.data = df_list
+        self.target = [torch.tensor(df[target].to_numpy(), dtype=dtype) for df in self.data]
+        self.features = [torch.tensor(df.drop([target], axis=1).to_numpy(), dtype=dtype) for df in self.data]
+
+    def __getitem__(self, index: int):
+        x = self.features[index]
+        y = self.target[index]
+        return x, y
+
+    def __len__(self) -> int:
+        return len(self.data)
+
+
+class ManyToOneDataset(Dataset):
+    """
+    A PyTorch Dataset for many-to-one data.
+
+    Args:
+        df_list (List[pd.DataFrame]): List of pandas DataFrames.
+        target (str): The target column name.
+        drop (Optional[Union[str, List[str]]]): Column(s) to drop from the DataFrames. Default is None.
+        dtype (torch.dtype): Data type for the tensors. Default is torch.float32.
+
+    Attributes:
+        data (List[pd.DataFrame]): List of pandas DataFrames with specified columns dropped.
+        target (List[torch.Tensor]): List of target tensors.
+        features (List[torch.Tensor]): List of feature tensors.
+
+    Examples:
+        >>> import pandas as pd
+        >>> from spotpython.data.manydataset import ManyToOneDataset
+        >>> df1 = pd.DataFrame({'feature1': [1, 2], 'feature2': [3, 4], 'target': [5, 6]})
+        >>> df2 = pd.DataFrame({'feature1': [7, 8], 'feature2': [9, 10], 'target': [11, 12]})
+        >>> dataset = ManyToOneDataset([df1, df2], target='target', drop='feature2')
+        >>> len(dataset)
+        2
+        >>> dataset[0]
+        (tensor([[1.],
+                 [2.]]), tensor(5.))
+    """
+
+    def __init__(
+        self,
+        df_list: List[pd.DataFrame],
+        target: str,
+        drop: Optional[Union[str, List[str]]] = None,
+        dtype: torch.dtype = torch.float32,
+    ):
+        try:
+            self.data = [df.drop(drop, axis=1) for df in df_list]
+        except KeyError:
+            self.data = df_list
+        self.target = [torch.tensor(df[target].to_numpy()[0], dtype=dtype) for df in self.data]
+        self.features = [torch.tensor(df.drop([target], axis=1).to_numpy(), dtype=dtype) for df in self.data]
+
+    def __getitem__(self, index: int):
+        x = self.features[index]
+        y = self.target[index]
+        return x, y
+
+    def __len__(self) -> int:
+        return len(self.data)