Dataset¶

class datacube.model.Dataset(product, metadata_doc, uris=None, uri=None, sources=None, indexed_by=None, indexed_time=None, archived_time=None, source_tree=None, derived_tree=None)[source]¶

A Dataset. A container of metadata, and refers typically to a multi-band raster on disk.

Most important parts are the metadata_doc and uri.

Dataset objects should be constructed by an index driver, or with the datacube.index.hl.Doc2Dataset

Parameters:

metadata_doc (Dict[str, Any]) – the document (typically a parsed json/yaml)
uris (Optional[List[str]]) – All active uris for the dataset

Members:

Attributes:

`bounds`	bounding box of the dataset in the native crs
`center_time`	mid-point of time range
`crs`	Return CRS if available
`extent`	valid extent of the dataset or None
`id`	UUID of a dataset
`indexed_by`	The User who indexed this dataset
`is_active`	Is this dataset active?
`is_archived`	Is this dataset archived?
`key_time`	datetime.datetime
`local_path`	A path to this dataset on the local filesystem (if available).
`local_uri`	Return the uri if it is local, or None.
`metadata_doc`	The document describing the dataset as a dictionary.
`sources`	The datasets that this dataset is derived from (if requested on load).
`uris`	List of active locations, newest to oldest.

Methods:

`has_multiple_uris`()	This is a 1.9-2.0 transitional method and will be removed in 2.0.
`legacy_uri`([schema])	This is a 1.9-2.0 transitional method and will be removed in 2.0.
`metadata_doc_without_lineage`()	Return metadata document without nested lineage datasets

property bounds: BoundingBox | None¶

bounding box of the dataset in the native crs

Type:: returns

center_time¶: mid-point of time range

property crs: CRS | None¶: Return CRS if available

extent¶

valid extent of the dataset or None

Type:: returns

has_multiple_uris()[source]¶

This is a 1.9-2.0 transitional method and will be removed in 2.0.

Returns true if the dataset has multiple locations.

Allows checking for multiple locations without tripping a deprecation warning.

Return type:: bool

property id: UUID¶: UUID of a dataset

indexed_by¶: The User who indexed this dataset

property is_active: bool¶

Is this dataset active?

(ie. dataset hasn’t been archived)

property is_archived: bool¶

Is this dataset archived?

(an archived dataset is one that is not intended to be used by users anymore: eg. it has been replaced by another dataset. It will not show up in search results, but still exists in the system via provenance chains or through id lookup.)

key_time¶

datetime.datetime

Type:: rtype

legacy_uri(schema=None)[source]¶

This is a 1.9-2.0 transitional method and will be removed in 2.0.

If the dataset has only one location, it returns that uri, but if the dataset has multiple locations, it calls various deprecated methods to achieve the legacy behaviour. It is intended for internal core use only.

Parameters:: schema (str | None)
Returns:

property local_path: Path | None¶: A path to this dataset on the local filesystem (if available).

property local_uri: str | None¶

Return the uri if it is local, or None.

Legacy behaviour: The latest local file uri, if any.

metadata_doc¶: The document describing the dataset as a dictionary. It is often serialised as YAML on disk or inside a NetCDF file, and as JSON-B inside the database index.

metadata_doc_without_lineage()[source]¶

Return metadata document without nested lineage datasets

Return type:: Dict[str, Any]

sources¶: The datasets that this dataset is derived from (if requested on load).

property uris: Sequence[str]¶

List of active locations, newest to oldest.

Multiple locations are deprecated, please use ‘uri’.