Initial Contact Detection (ICD)

Algorithms to detect ICs within raw IMU data during a gait sequence.

Algorithms

IcdShinImproved([axis])	Detect initial contacts using the Shin [R2799e68cb60b-1] algorithm, with improvements by Ionescu et al. [R2799e68cb60b-2].
IcdIonescu(*[, pre_filter, cwt_width])	Implementation of the IC detection algorithm by McCamley (2012) [R44ff74c6888c-1] modified by Ionescu (2020) [R44ff74c6888c-2].
IcdHKLeeImproved([axis])	Detect initial contacts using the HKLee [R036c78852f60-1] algorithm, with improvements by Ionescu et al. [R036c78852f60-2].

Pipelines

pipeline.IcdEmulationPipeline(algo, *[, ...])

Run an ICD algorithm in isolation on a Gait Dataset.

Utils

refine_gs(ic_list)

Get the start and end values of a refined gait sequence that starts at the first and ends at the last IC.

Evaluation

calculate_matched_icd_performance_metrics(...)	Calculate performance metrics for initial contact detection results.
calculate_true_positive_icd_error(...[, groupby])	Calculate error metrics for initial contact detection results.
categorize_ic_list(*, ic_list_detected, ...)	Evaluate an initial contact list against a reference contact-by-contact.
get_matching_ics(*, metrics_detected, ...)	Extract the detected and reference initial contacts that are considered as matches sequence-by-sequence (tps).

Evaluation Scores

These scores are expected to be used in combination with Evaluation and EvaluationCV or directly with cross_validation and validation.

icd_score

Scorer for ICD algorithms.

icd_per_datapoint_score(pipeline, datapoint)	Evaluate the performance of an ICD algorithm on a single datapoint.
icd_final_agg(agg_results, single_results, ...)	Aggregate the performance metrics of an ICD algorithm over multiple datapoints.

Base Classes

Base class for ICs detectors.

BaseIcDetector()

Base class for IC-detectors.

Docu-helper

base_icd_docfiller(func)

Decorator to fill common parts of the docstring for subclasses of BaseIcdDetector.