Dataset

class anemoi.datasets.usage.dataset.Dataset

dates_interval_to_indices(start: None | str | datetime, end: None | str | datetime) → list[int]

Convert date interval to a list of indices.

Parameters:

• start (None, str, or datetime.datetime) – The start date.

• end (None, str, or datetime.datetime) – The end date.

Returns:

The list of indices.

Return type:

list of int

provenance() → dict[str, Any]

Return the provenance information of the dataset.

Returns:

The provenance information.

Return type:

dict

property typed_variables: dict[str, Any]

Return the variables with their types.

metadata() → dict[str, Any]

Return the metadata of the dataset.

Returns:

The metadata.

Return type:

dict

property start_date: datetime64

Return the start date of the dataset.

property end_date: datetime64

Return the end date of the dataset.

supporting_arrays() → dict[str, ndarray[tuple[Any, ...], dtype[Any]]]

Return the supporting arrays to be saved in the checkpoints.

Returns:

The supporting arrays.

Return type:

dict

property grids: tuple[int | slice | EllipsisType, ...]

Return the grid shape of the dataset.

abstract property variables: list[str]

Return the list of variables in the dataset.

abstract property frequency: timedelta

Return the frequency of the dataset.

abstract property dates: ndarray[tuple[Any, ...], dtype[datetime64]]

Return the dates in the dataset.

abstract property name_to_index: dict[str, int]

Return the mapping of variable names to indices.

abstract property shape: tuple[int, ...]

Return the shape of the dataset.

abstract property field_shape: tuple[int, ...]

Return the shape of the fields in the dataset.

abstract property dtype: dtype

Return the data type of the dataset.

abstract property latitudes: ndarray[tuple[Any, ...], dtype[Any]]

Return the latitudes in the dataset.

abstract property longitudes: ndarray[tuple[Any, ...], dtype[Any]]

Return the longitudes in the dataset.

abstract property missing: set[int]

Return the set of missing indices in the dataset.

abstract property statistics: dict[str, ndarray[tuple[Any, ...], dtype[Any]]]

Return the statistics of the dataset.

abstractmethod statistics_tendencies(delta: timedelta | None = None) → dict[str, ndarray[tuple[Any, ...], dtype[Any]]]

Return the tendencies of the statistics in the dataset.

Parameters:

delta (datetime.timedelta, optional) – The time delta for computing tendencies.

Returns:

The tendencies.

Return type:

dict