MultiSampleAnalysis.extract_features

Module: workflow.py

Builds event-level feature table including baseline, morphology and statistical descriptors.

Parameters

• custom_feature_fns (dict[str, Callable] | None): Optional custom feature functions; each receives event segment and returns dict.
• max_event_per_sample (int | None): Maximum number of events to extract per sample (take the first N events). None means extract all events.

Returns

• pandas.DataFrame (feature table).

Notes

• Shows sample-level and event-level progress bars automatically when tqdm is available in the environment.
• Direction-aware feature convention:
• detect_direction="down": delta_i = global_baseline - segment_mean, blockade_ratio = delta_i / global_baseline.
• detect_direction="up": delta_i = -global_baseline - (-segment_mean), blockade_ratio = delta_i / (-global_baseline).