AFL.double_agent.PipelineOp module

class AFL.double_agent.PipelineOp.PipelineOp(name: str | None | List[str] = None, input_variable: str | None | List[str] = None, output_variable: str | None | List[str] = None, input_prefix: str | None | List[str] = None, output_prefix: str | None | List[str] = None)

Bases: ABC

Abstract base class for data processors. All operations in AFL.double_agent should inherit PipelineOpBase.

Parameters:

• name (Optional[str] | List[str]) – The name to use when added to a Pipeline. This name is used when calling Pipeline.search()

• input_variable (Optional[str] | List[str]) – The name of the xarray.Dataset data variable to extract from the input dataset

• output_variable (Optional[str] | List[str]) – The name of the variable to be inserted into the xarray.Dataset by this PipelineOp

• input_prefix (Optional[str] | List[str]) – Prefix for input variables when using pattern matching

• output_prefix (Optional[str] | List[str]) – Prefix for output variables when using pattern matching

PROVIDED_KEYS: tuple = ()

add_to_dataset(dataset, copy_dataset=True)

Adds (xarray) data in output dictionary to provided xarray dataset

add_to_tiled(tiled_data)

Adds data in output dictionary to provided tiled catalogue

abstract calculate(dataset: Dataset) → Self

copy() → Self

classmethod from_json(json_data, objects: Dict[str, object] | None = None)

Rebuild an op from to_json() output.

Parameters:

• json_data (dict) – {"class": ..., "args": {...}}.

• objects (dict, optional) – Live objects for {"ref": ...} arguments, keyed by reference ("<op name>.<arg>"), e.g. pipeline.objects of the pipeline that was exported or AFL.double_agent.Pipeline.load_environment() of a module that defines it.

Raises:

UnresolvedReferenceError – If an argument is a reference and objects does not contain it.

live_objects() → Dict[str, object]

Constructor arguments kept in memory (no JSON form), keyed by their export reference.

plot(sample_dim: str = 'sample', **mpl_kwargs) → Figure

Plots the output of the PipelineOp.

This method attempts to guess how to plot the data produced by the operation.

to_json()

Serialize the class path and constructor arguments to a JSON-safe dict.

Arguments without a JSON form are written as {"ref": "<name>.<arg>"}; the live objects are available from live_objects() and must be passed back to from_json() (objects=) when the pipeline is reloaded.