nicetoolbox.detectors.base_detector.BaseDetector¶

class nicetoolbox.detectors.base_detector.BaseDetector(io: SequenceIO, data: SequenceData, subsequence_context: SubsequenceContext, algorithm_instance: str)[source]¶

Bases: ABC

Abstract base class for ALL detectors.

Defines the common interface that both method and feature detectors implement. This enables a unified detector loop in main.py.

Initialize base detector with references.

Subclasses should call super().__init__() and set inference_config.

Methods

`compute_output_folders`	Compute extra output folders for all components.
`compute_result_folders`	Compute result folders for all components.
`compute_viz_folders`	Compute visualization folders for all components.
`resolve_inputs`	Return the inputs to resolve for this detector.
`resolve_outputs`	Return the outputs this detector declares it will produce.
`run`	Execute the detector's main computation.
`visualization`	Visualize detector output.

Attributes

`inputs`
`outputs`
`predictions_mapping`	Access predictions mapping from runtime config.
`algorithm_type`
`algorithm_instance`
`data`
`io`
`subsequence_context`
`detector_config`
`inference_config`
`components`
`loaded_inputs`
`declared_outputs`
`visualize`

compute_output_folders(requires_out_folder: bool) → Dict[str, str][source]¶: Compute extra output folders for all components.

compute_result_folders() → Dict[str, str][source]¶: Compute result folders for all components.

compute_viz_folders(visualize: bool) → Dict[str, str][source]¶: Compute visualization folders for all components.

property predictions_mapping¶: Access predictions mapping from runtime config.

Return the inputs to resolve for this detector.

Defaults to the class-level inputs. Override to vary inputs by config (e.g. add an optional input only when a flag is set).

Return the outputs this detector declares it will produce.

Defaults to the class-level outputs. Override to vary outputs by config (e.g. pick the npz_key/schema based on a per-instance dimension).

abstract run() → Any | None[source]¶

Execute the detector’s main computation.

For method detectors: runs subprocess inference + post_inference() For feature detectors: runs compute()

Returns:: Optional data for visualization (feature detectors currently return computed data)

abstract visualization(data: Any) → None[source]¶

Visualize detector output.