PrerequisiteInput

class lsst.pipe.base.connectionTypes.PrerequisiteInput(name: str, storageClass: str, doc: str = '', multiple: bool = False, _deprecation_context: str = '', dimensions: Iterable[str] = (), isCalibration: bool = False, deferLoad: bool = False, minimum: int = 1, lookupFunction: Callable[[DatasetType, Registry, DataCoordinate, Sequence[str]], Iterable[DatasetRef]] | None = None, *, deprecated: str | None = None)

Bases: BaseInput

Class used for declaring PipelineTask prerequisite connections.

Attributes:

namestr

The default name used to identify the dataset type.

storageClassstr

The storage class used when (un)/persisting the dataset type.

multiplebool

Indicates if this connection should expect to contain multiple objects of the given dataset type. Tasks with more than one connection with multiple=True with the same dimensions may want to implement PipelineTaskConnections.adjustQuantum to ensure those datasets are consistent (i.e. zip-iterable) in PipelineTask.runQuantum and notify the execution system as early as possible of outputs that will not be produced because the corresponding input is missing.

dimensionsiterable of str

The lsst.daf.butler.Butler lsst.daf.butler.Registry dimensions used to identify the dataset type identified by the specified name.

minimumbool

Minimum number of datasets required for this connection, per quantum. This is checked in the base implementation of PipelineTaskConnections.adjustQuantum, which raises FileNotFoundError (causing QuantumGraph generation to fail). PipelineTask implementations may provide custom adjustQuantum implementations for more fine-grained or configuration-driven constraints, as long as they are compatible with this minium.

lookupFunctiontyping.Callable, optional

An optional callable function that will look up PrerequisiteInputs using the DatasetType, registry, quantum dataId, and input collections passed to it. If no function is specified, the default temporal spatial lookup will be used.

Raises:

TypeError

Raised if minimum is greater than one but multiple=False.

Notes

Prerequisite inputs are used for datasets that must exist in the data repository before a pipeline including this is run; they cannot be produced by another task in the same pipeline.

In exchange for this limitation, they have a number of advantages relative to regular Input connections:

• The query used to find them then during QuantumGraph generation can be fully customized by providing a lookupFunction.

• Failed searches for prerequisites during QuantumGraph generation will usually generate more helpful diagnostics than those for regular Input connections.

• The default query for prerequisite inputs relates the quantum dimensions directly to the dimensions of its dataset type, without being constrained by any of the other dimensions in the pipeline. This allows them to be used for temporal calibration lookups (which regular Input connections cannot do at present) and to work around QuantumGraph generation limitations involving cases where naive spatial overlap relationships between dimensions are not desired (e.g. a task that wants all detectors in each visit for which the visit overlaps a tract, not just those where that detector+visit combination overlaps the tract).

• Prerequisite inputs may be optional (regular inputs are never optional).

Attributes Summary

deferLoad
deprecated
dimensions
doc
isCalibration
lookupFunction
minimum
multiple

Attributes Documentation

deferLoad: bool = False

deprecated: str | None = None

dimensions: Iterable[str] = ()

doc: str = ''

isCalibration: bool = False

lookupFunction: Callable[[DatasetType, Registry, DataCoordinate, Sequence[str]], Iterable[DatasetRef]] | None = None

minimum: int = 1

multiple: bool = False