BasicMotions

Task: Classification Classes: 4 Channels: 6 Length: 100

Overview

Multivariate time series classification dataset from the UEA archive (Bagnall et al., 2018). It was recorded in a student project: four students performed four activities five times each while wearing a smart watch, which recorded the x, y and z axes of its accelerometer and its gyroscope (6 channels) at 10 Hz for 10 seconds.

../../_images/basic_motions_teaser.png

The archive does not document which channels belong to the accelerometer and which to the gyroscope, so the channels are kept in the archive’s order. The loader passes the archive’s data through unchanged.

  • Train: 40 recordings (10 per class).

  • Test: 40 recordings (10 per class).

Class names are the archive’s labels in lower case: badminton, running, standing, walking. (The archive’s description calls standing “resting”.)

Data Structure

Key

Type

Description

series

list of list of float32

Smart watch recording, shape (100, 6): time steps by channels

label

int

Activity class (0-3)

Usage Example

import numpy as np
from stable_datasets.timeseries import BasicMotions

train = BasicMotions(split="train")
sample = train[0]
x = np.asarray(sample["series"])  # (100, 6)
print(x.shape, train.features["label"].names[sample["label"]])

References

Citation

@article{bagnall2018uea,
  title={The UEA multivariate time series classification archive, 2018},
  author={Bagnall, Anthony and Dau, Hoang Anh and Lines, Jason and Flynn, Michael and Large, James and Bostrom, Aaron and Southam, Paul and Keogh, Eamonn},
  journal={arXiv preprint arXiv:1811.00075},
  year={2018}
}