kinoml.datasets.core ==================== .. py:module:: kinoml.datasets.core .. autoapi-nested-parse:: Base classes for ``DatasetProvider``-like objects Module Contents --------------- .. py:data:: logger .. py:class:: BaseDatasetProvider Bases: :py:obj:`object` API specification for dataset providers .. py:method:: from_source(path_or_url=None, **kwargs) :classmethod: :abstractmethod: Parse CSV/raw files to object model. .. py:method:: observation_model(backend='pytorch') :abstractmethod: .. py:property:: systems :abstractmethod: .. py:property:: measurement_type :abstractmethod: .. py:method:: measurements_as_array(reduce=np.mean) :abstractmethod: .. py:method:: measurements_by_group() :abstractmethod: .. py:property:: conditions :abstractmethod: .. py:method:: featurize(*featurizers: Iterable[kinoml.features.core.BaseFeaturizer]) :abstractmethod: .. py:method:: featurized_systems(key='last') :abstractmethod: .. py:method:: to_dataframe(*args, **kwargs) :abstractmethod: .. py:method:: to_pytorch(**kwargs) :abstractmethod: .. py:method:: to_tensorflow(*args, **kwargs) :abstractmethod: .. py:method:: to_numpy(*args, **kwargs) :abstractmethod: .. py:class:: DatasetProvider(measurements: Iterable[kinoml.core.measurements.BaseMeasurement], metadata: dict = None) Bases: :py:obj:`BaseDatasetProvider` Base object for all DatasetProvider classes. :param measurements: A DatasetProvider holds a list of ``kinoml.core.measurements.BaseMeasurement`` objects (or any of its subclasses). They must be of the same type! :type measurements: list of BaseMeasurement :param metadata: Extra information for provenance. :type metadata: dict .. note:: All measurements must be of the same type! If they are not, consider using ``MultiDatasetProvider`` instead. .. py:attribute:: _raw_data :value: None .. py:attribute:: measurements .. py:attribute:: metadata .. py:method:: __len__() .. py:method:: __getitem__(subscript) .. py:method:: __repr__() -> str .. py:method:: from_source(path_or_url=None, **kwargs) :classmethod: :abstractmethod: Parse CSV/raw file to object model. This method is responsible of generating the objects for ``self.measurements``, if relevant. Additional kwargs will be passed to ``__init__``. You must define this in your subclass. .. py:method:: featurize(featurizer: kinoml.features.core.BaseFeaturizer) Given a collection of ``kinoml.features.core.BaseFeaturizers``, apply them to the systems present in the ``self.measurements``. :param featurizer: Featurization scheme that will be applied to the systems, in a stacked way. :type featurizer: BaseFeaturizer .. note:: TODO: * Will the systems be properly featurized with Dask? .. py:method:: _post_featurize(featurizer: kinoml.features.core.BaseFeaturizer) Remove measurements with systems, that were not successfully featurized. :param featurizer: The used featurizer. :type featurizer: BaseFeaturizer .. py:method:: featurized_systems(key='last', clear_after=False) Return the ``key`` featurized objects from all systems. .. py:method:: _to_dataset(style='pytorch') :abstractmethod: Generate a clean