datatime package

Submodules

datatime.classes module

class datatime.classes.TimeSeriesClassificationDataset(X_train: Array, y_train: ndarray[Any, dtype[Any]], X_test: Array, y_test: ndarray[Any, dtype[Any]], labels: Optional[Dict[Any, Any]] = None, name: str = '')

Bases: TimeSeriesDataset

map_labels(y: ndarray[Any, dtype[Any]]) → Any

class datatime.classes.TimeSeriesDataset

Bases: object

class datatime.classes.TimeSeriesForecastingDataset(X: Array, Y: Array, name: str = '')

Bases: TimeSeriesDataset

XY() → Array

class datatime.classes.TimeSeriesMultioutputDataset(X_train: Array, Y_train: DataFrame, X_test: Array, Y_test: DataFrame, name: str = '')

Bases: TimeSeriesDataset

X() → Array

XY() → Tuple[Array, DataFrame]

Y() → DataFrame

class datatime.classes.TimeSeriesRegressionDataset(X_train: Array, y_train: ndarray[Any, dtype[Any]], X_test: Array, y_test: ndarray[Any, dtype[Any]], name: str = '')

Bases: TimeSeriesDataset

datatime.conversion module

datatime.conversion.awkward_to_flat(X: Array) → Array

datatime.conversion.awkward_to_pyts(X: Array) → ndarray[Any, dtype[Any]]

datatime.conversion.awkward_to_sktime(X: Array) → DataFrame

datatime.conversion.awkward_to_tslearn(X: Array) → ndarray[Any, dtype[Any]]

datatime.conversion.has_equal_length_signals(X: Array) → bool

datatime.conversion.has_same_number_of_signals(X: Array) → bool

datatime.conversion.is_multivariate(X: Array) → bool

datatime.conversion.is_univariate(X: Array) → bool

datatime.conversion.pyts_to_awkward(X: ndarray[Any, dtype[ScalarType]]) → Array

datatime.conversion.sktime_to_awkward(X: DataFrame) → Array

datatime.conversion.tslearn_to_awkward(X: ndarray[Any, dtype[ScalarType]]) → Array

datatime.database_utils module

datatime.database_utils.X_info(X: Array) → Tuple[int, int, int, int, bool, bool]

datatime.database_utils.cached_datasets_dict(root: Optional[Path] = None) → Dict[str, List[str]]

datatime.database_utils.dataset_info(dataset: Union[TimeSeriesClassificationDataset, TimeSeriesRegressionDataset, TimeSeriesForecastingDataset]) → Dict[str, Any]

datatime.database_utils.datasets_info(names: List[str]) → DataFrame

datatime.database_utils.datasets_list(tasks: Optional[List[str]] = None)

datatime.database_utils.datasets_table(tasks: Optional[List[str]] = None) → DataFrame

datatime.database_utils.is_cached(dataset_name: str, task: str) → bool

datatime.database_utils.load_classification_dataset(name: str, nan_value: float = nan, origin: str = 'gdrive', path: Optional[str] = None) → Tuple[Array, ndarray[Any, dtype[Any]], Array, ndarray[Any, dtype[Any]], Dict[int, str]]

datatime.database_utils.load_dataset(name: str, nan_value: float = nan) → Union[TimeSeriesClassificationDataset, TimeSeriesRegressionDataset, TimeSeriesForecastingDataset, TimeSeriesMultioutputDataset]

Load a time series dataset.

Parameters

• name – The name of the dataset to load.

• nan_value – The value that represents a missing value.

Returns

A TimeSeriesDataset object.

datatime.database_utils.load_forecasting_dataset(name: str, nan_value: float = nan, origin='gdrive') → Tuple[Array, Array]

datatime.database_utils.load_multioutput_dataset(name: str, nan_value: float = nan, origin='gdrive', path: Optional[str] = None, load_train: bool = True, load_test: bool = True) → Tuple[Optional[Array], DataFrame, Optional[Array], DataFrame]

datatime.database_utils.load_regression_dataset(name: str, nan_value: float = nan, origin='gdrive') → Tuple[Array, ndarray[Any, dtype[Any]], Array, ndarray[Any, dtype[Any]]]

datatime.download_utils module

datatime.download_utils.download_dataset(name: str, origin: str) → None

datatime.gdrive_utils module

datatime.gdrive_utils.download_file_from_google_drive(id: str, destination: str) → None

datatime.gdrive_utils.get_id_to_download_gdrive(dataset_name: str) → DataFrame

datatime.github_utils module

datatime.github_utils.download_raw_file_from_github(url: str) → str

datatime.github_utils.get_id_to_download_github(dataset_name: str) → DataFrame

datatime.github_utils.get_raw_file_dataset_url_github(dataset_name: str, task: str, prefix: str = 'https://raw.githubusercontent.com/fspinna/datatime/main/datatime/datasets') → str

datatime.utils module

datatime.utils.fill_none(*args: Array, replace_with: float = nan) → List[Array]

datatime.utils.get_default_dataset_path(dataset_name: str, task: str) → Path

datatime.utils.get_project_root() → Path

datatime.utils.map_labels(y: ndarray[Any, dtype[Any]], labels: Dict[Any, Any]) → Any

datatime.utils.shape(X: Array) → Tuple

Module contents