Features

class gaitalytics.features.CycleFeaturesCalculation(config: MappingConfigs, **kwargs)

abstract _calculate(trial: Trial) → DataArray

Calculate the features for a trial.

Parameters:: trial – The trial for which to calculate the features.
Returns:: An xarray DataArray containing the calculated features.

static _flatten_features(features: DataArray) → DataArray

Flatten the features into a single dimension.

The dimension channel will be integrated into the features dim

Parameters:: features – The features to be flattened.
Returns:: The reshaped features.

calculate(trial: TrialCycles) → DataArray

Calculate the features for a trial.

Calls the _calculate method for each cycle in the trial and combines results into a single DataArray.

Parameters:: trial – The trial for which to calculate the features.
Returns:: An xarray DataArray containing the calculated features.

static get_event_times(trial_events: DataFrame | None) → tuple[float, float, float, float, float]

Checks the sequence of events in the trial and returns the times.

Parameters:: trial_events – The events to be checked and extracted.
Returns:: The times of the events. in following order [contra_fo, contra_fs, ipsi_fo, end_time]
Raises:: ValueError – If the sequence of events is incorrect.

class gaitalytics.features.FeatureCalculation(config: MappingConfigs, **kwargs)

Base class for feature calculations.

This class provides a common interface for calculating features.

__init__(config: MappingConfigs, **kwargs)

Initializes a new instance of the BaseFeatureCalculation class.

Parameters:

config – The mapping configuration to use for the feature calculation.
**kwargs – Currently not used.

abstract calculate(trial: TrialCycles) → DataArray

Calculate the features for a trial.

Parameters:: trial – The trial for which to calculate the features.
Returns:: An xarray DataArray containing the calculated features.

class gaitalytics.features.PhaseTimeSeriesFeatures(config: MappingConfigs, **kwargs)

Calculate phase time series features for a trial.

This class calculates following phase time series features for a trial.

stand_min
stand_max
stand_mean
stand_median
stand_std
stand_amplitude
swing_min
swing_max
swing_mean
swing_median
swing_std
swing_amplitude

class gaitalytics.features.PointDependentFeature(config: MappingConfigs, **kwargs)

_get_marker_data(trial: Trial, marker: MappedMarkers) → DataArray

Get the marker data for a trial.

Parameters:

trial – The trial for which to get the marker data.
marker – The marker to get the data for.

Returns:

An xarray DataArray containing the marker data.

_get_progression_vector(trial: Trial) → DataArray

Calculate the progression vector for a trial.

The progression vector is the vector from the sacrum to the anterior hip marker.

Parameters:: trial – The trial for which to calculate the progression vector.
Returns:: An xarray DataArray containing the calculated progression vector.

_get_sacrum_marker(trial: Trial) → DataArray

Get the sacrum marker data for a trial.

Try to get the sacrum marker data from the trial. If the sacrum marker not found calculate from posterior hip markers

Parameters:: trial – The trial for which to get the marker data.
Returns:: An xarray DataArray containing the sacrum marker data.

class gaitalytics.features.SpatialFeatures(config: MappingConfigs, **kwargs)

Calculate spatial features for a trial.

This class calculates following spatial features for a trial. - step_length - step_width - minimal_toe_clearance - AP_margin_of_stability - AP_base_of_support - AP_xcom - ML_margin_of_stability - ML_base_of_support - ML_xcom

class gaitalytics.features.TemporalFeatures(config: MappingConfigs, **kwargs)

Calculate temporal features for a trial.

This class calculates following temporal features for a trial.

double_support
single_support
stance_duration_prec
swing_duration_prec
opposite_foot_off_prec
opposite_foot_contact_prec
stride_duration
stance_duration
cadence

class gaitalytics.features.TimeSeriesFeatures(config: MappingConfigs, **kwargs)

Calculate time series features for a trial.

This class calculates following time series features for a trial. - min - max - mean - median - std