hopeit.dataframes 0.25.0b6__tar.gz → 0.25.0b9__tar.gz

{hopeit_dataframes-0.25.0b6 → hopeit_dataframes-0.25.0b9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: hopeit.dataframes
-Version: 0.25.0b6
+Version: 0.25.0b9
 Summary: Hopeit Engine Dataframes Toolkit
 Home-page: https://github.com/hopeit-git/hopeit.engine
 Author: Leo Smerling and Pablo Canto
@@ -13,7 +13,6 @@ Project-URL: GitHub: repo, https://github.com/hopeit-git/hopeit.engine
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Intended Audience :: Developers
 Classifier: Programming Language :: Python
-Classifier: Programming Language :: Python :: 3.8
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
@@ -24,9 +23,9 @@ Classifier: Operating System :: Microsoft :: Windows
 Classifier: Topic :: Internet :: WWW/HTTP
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Classifier: Framework :: AsyncIO
-Requires-Python: >=3.8
+Requires-Python: >=3.9
 Description-Content-Type: text/markdown
-Requires-Dist: hopeit.engine[fs-storage]==0.25.0b6
+Requires-Dist: hopeit.engine[fs-storage]==0.25.0b9
 Requires-Dist: pandas
 Requires-Dist: numpy
 Provides-Extra: pyarrow

{hopeit_dataframes-0.25.0b6 → hopeit_dataframes-0.25.0b9}/setup.py

@@ -18,7 +18,6 @@ setuptools.setup(
         "License :: OSI Approved :: Apache Software License",
         "Intended Audience :: Developers",
         "Programming Language :: Python",
-        "Programming Language :: Python :: 3.8",
         "Programming Language :: Python :: 3.9",
         "Programming Language :: Python :: 3.10",
         "Programming Language :: Python :: 3.11",
@@ -46,7 +45,7 @@ setuptools.setup(
     package_data={
         "hopeit.dataframes": ["py.typed"],
     },
-    python_requires=">=3.8",
+    python_requires=">=3.9",
     install_requires=[
         f"hopeit.engine[fs-storage]=={version['ENGINE_VERSION']}",
         "pandas",

{hopeit_dataframes-0.25.0b6 → hopeit_dataframes-0.25.0b9}/src/hopeit/dataframes/dataframe.py

@@ -1,8 +1,10 @@
 """
 DataFrames type abstractions.
 """
+
 import dataclasses
 from datetime import date, datetime, timezone
+from functools import partial
 from typing import Any, Callable, Dict, Generic, Iterator, List, Type, TypeVar
 
 import numpy as np
@@ -23,11 +25,27 @@ DataFrameT = TypeVar("DataFrameT")
 
 
 @dataclasses.dataclass
-class DataFrameMetadata():
+class DataFrameMetadata:
     columns: List[str]
     fields: Dict[str, FieldInfo]
 
 
+# Functions to do type coercion
+def _series_to_int(x: pd.Series) -> pd.Series:
+    return x.astype(np.int64)
+
+
+def _series_to_float(x: pd.Series) -> pd.Series:
+    return x.astype(np.float64)
+
+
+def _series_to_str(x: pd.Series) -> pd.Series:
+    return x.astype(str)
+
+
+_series_to_utc_datetime = partial(pd.to_datetime, utc=True)
+
+
 class DataFrameMixin(Generic[DataFrameT, DataObject]):
     """
     MixIn class to add functionality for DataFrames dataobjects
@@ -36,11 +54,11 @@ class DataFrameMixin(Generic[DataFrameT, DataObject]):
     """
 
     DATATYPE_MAPPING = {
-        int: lambda x: x.astype(np.int64),
-        float: lambda x: x.astype(np.float64),
-        str: lambda x: x.astype(str),
+        int: _series_to_int,
+        float: _series_to_float,
+        str: _series_to_str,
         date: pd.to_datetime,
-        datetime: pd.to_datetime,
+        datetime: _series_to_utc_datetime,
     }
 
     def __init__(self) -> None:
@@ -51,9 +69,7 @@ class DataFrameMixin(Generic[DataFrameT, DataObject]):
         raise NotImplementedError  # must use @dataframe decorator  # pragma: no cover
 
     @staticmethod
-    def __init_from_series__(
-        self, **series: pd.Series
-    ):  # pylint: disable=bad-staticmethod-argument
+    def __init_from_series__(self, **series: pd.Series):  # pylint: disable=bad-staticmethod-argument
         df = pd.DataFrame(series)
         df.index.name = None  # Removes index name to avoid colisions with series name
         if self.__data_object__["validate"]:
@@ -89,10 +105,7 @@ class DataFrameMixin(Generic[DataFrameT, DataObject]):
         return self._from_df(self.__df[key])
 
     def _to_dataobjects(self) -> List[DataObject]:
-        return [
-            self.DataObject(**fields)
-            for fields in self.__df.to_dict(orient="records")
-        ]
+        return [self.DataObject(**fields) for fields in self.__df.to_dict(orient="records")]
 
     def event_id(self, *args, **kwargs) -> str:
         return ""
@@ -136,7 +149,7 @@ def dataframe(
         if hasattr(cls, "__annotations__") and hasattr(cls, "__dataclass_fields__"):
             amended_class = type(
                 cls.__name__,
-                (DataFrameMixin, ) + cls.__mro__,
+                (DataFrameMixin,) + cls.__mro__,
                 dict(cls.__dict__),
             )
             setattr(amended_class, "__init__", DataFrameMixin.__init_from_series__)
@@ -145,7 +158,7 @@ def dataframe(
 
     def add_dataframe_metadata(cls):
         serialized_fields = {k: (v.annotation, v) for k, v in fields(cls).items()}
-        dataobject_type = create_model(cls.__name__+"DataObject", **serialized_fields)
+        dataobject_type = create_model(cls.__name__ + "DataObject", **serialized_fields)
         dataobject_type = dataobject(dataobject_type, unsafe=True)
 
         setattr(cls, "DataObject", dataobject_type)

{hopeit_dataframes-0.25.0b6 → hopeit_dataframes-0.25.0b9}/src/hopeit/dataframes/serialization/dataset.py

@@ -1,5 +1,4 @@
-"""Dataset objects definition, used as a result of serialized dataframes
-"""
+"""Dataset objects definition, used as a result of serialized dataframes"""
 
 from importlib import import_module
 from typing import Generic, Type, TypeVar
@@ -13,6 +12,7 @@ DataFrameT = TypeVar("DataFrameT")
 @dataclass
 class Dataset(Generic[DataFrameT]):
     """Persisted representation of a @dataframe object"""
+
     protocol: str
     partition_key: str
     key: str

{hopeit_dataframes-0.25.0b6 → hopeit_dataframes-0.25.0b9}/src/hopeit/dataframes/serialization/files.py

@@ -1,5 +1,4 @@
-"""Support for `@dataframes` serialization to files
-"""
+"""Support for `@dataframes` serialization to files"""
 
 import io
 from importlib import import_module
@@ -59,9 +58,7 @@ class DatasetFileStorage(Generic[DataFrameT]):
     async def load(self, dataset: Dataset) -> EventPayloadType:
         """Loads @dataframe annotated object using Dataset metadata"""
         datatype: Type[DataFrameT] = find_dataframe_type(dataset.datatype)
-        data = await self.storage.get_file(
-            dataset.key, partition_key=dataset.partition_key
-        )
+        data = await self.storage.get_file(dataset.key, partition_key=dataset.partition_key)
         if data is None:
             raise FileNotFoundError(dataset.key)
         df = pd.read_parquet(io.BytesIO(data), engine="pyarrow")

{hopeit_dataframes-0.25.0b6 → hopeit_dataframes-0.25.0b9}/src/hopeit/dataframes/serialization/settings.py

@@ -1,5 +1,4 @@
-"""Support for plugin configuration
-"""
+"""Support for plugin configuration"""
 
 from typing import Optional
 

{hopeit_dataframes-0.25.0b6 → hopeit_dataframes-0.25.0b9}/src/hopeit.dataframes.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: hopeit.dataframes
-Version: 0.25.0b6
+Version: 0.25.0b9
 Summary: Hopeit Engine Dataframes Toolkit
 Home-page: https://github.com/hopeit-git/hopeit.engine
 Author: Leo Smerling and Pablo Canto
@@ -13,7 +13,6 @@ Project-URL: GitHub: repo, https://github.com/hopeit-git/hopeit.engine
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Intended Audience :: Developers
 Classifier: Programming Language :: Python
-Classifier: Programming Language :: Python :: 3.8
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
@@ -24,9 +23,9 @@ Classifier: Operating System :: Microsoft :: Windows
 Classifier: Topic :: Internet :: WWW/HTTP
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Classifier: Framework :: AsyncIO
-Requires-Python: >=3.8
+Requires-Python: >=3.9
 Description-Content-Type: text/markdown
-Requires-Dist: hopeit.engine[fs-storage]==0.25.0b6
+Requires-Dist: hopeit.engine[fs-storage]==0.25.0b9
 Requires-Dist: pandas
 Requires-Dist: numpy
 Provides-Extra: pyarrow

hopeit_dataframes-0.25.0b9/src/hopeit.dataframes.egg-info/requires.txt

@@ -0,0 +1,6 @@
+hopeit.engine[fs-storage]==0.25.0b9
+pandas
+numpy
+
+[pyarrow]
+pyarrow

hopeit_dataframes-0.25.0b6/src/hopeit.dataframes.egg-info/requires.txt

@@ -1,6 +0,0 @@
-hopeit.engine[fs-storage]==0.25.0b6
-pandas
-numpy
-
-[pyarrow]
-pyarrow