biopsykit.signals.imu.wear_detection module

Module for detection non-wear times from raw acceleration signals.

class biopsykit.signals.imu.wear_detection.WearDetection(sampling_rate)

Bases: object

Initialize a new WearDetection instance.

Parameters

sampling_rate (float) – sampling rate of recorded data in Hz

sampling_rate: float

predict(data)

Predict non-wear times from acceleration data.

Parameters

data (array_like) – input acceleration data. Must be 3-d.

Returns

dataframe with wear (1) and non-wear (0) times per 15 minute interval

Return type

DataFrame

static get_major_wear_block(data)

Return major wear block.

The major wear block is the longest continuous wear block in the data.

Parameters

data (DataFrame) – data with wear detection applied. The dataframe is expected to have a “wear” column.

Returns

• start (datetime or int) – start of major wear block as datetime or int index

• end (datetime or int) – end of major wear block as datetime or int index

Return type

tuple[datetime.datetime | int, datetime.datetime | int]

See also

predict()

apply wear detection on accelerometer data

static cut_to_wear_block(data, wear_block)

Cut data to wear block.

Parameters

• data (DataFrame) – input data that contains wear block

• wear_block (tuple) – tuple with start and end times of wear block. The type of wear_block depends on the index of data. (datetime or int)

Returns

data cut to wear block

Return type

DataFrame

biopsykit.signals.imu.static_moment_detection module biopsykit.signals.pep package