Preprocessed accelerometer and ground reaction force data from countermovement jumps (Python format)

Preprocessed triaxial accelerometer and vertical ground reaction force (vGRF) data from countermovement jumps (CMJs), provided as Python-readable NumPy .npz archives. These data derive from the accelerometer dataset in White et al. (2022, PLOS ONE) but with additional quality exclusions and preprocessing applied, reducing the dataset from 69 to 67 participants (663 trials from 691) after removal of trials exhibiting ADC clipping (12 jumps from 2 participants) and likely sensor miscalibration (16 jumps from one participant). Participants were healthy sports science students, free of injury, all of whom had given their prior written consent. Ethical approval was given by the relevant Swansea University ethics committee, which included further analysis of the data. 45 males, 22 females (21.6 ± 1.5 yrs; standing height 1.74 ± 0.10 m; body mass 73.1 ± 13.1 kg) Trigno sensor (Delsys Inc, MA, USA), 250 Hz, attached to the lower back (L5) Two portable force platforms (Kistler, Winterthur, Switzerland), 1000 Hz vGRF signals have been downsampled to 250 Hz, normalised by body weight, and aligned to the instant of takeoff. A 2,000 ms pre-takeoff window (500 samples) has been extracted from both signals. Three dataset files are provided: cmj_dataset_noarms.npz — jumps without arm swing (331 trials) cmj_dataset_arms.npz — jumps with arm swing (332 trials) cmj_dataset_both.npz — both conditions combined (663 trials) Each file contains: accelerometer signals (3-axis, in g), vGRF signals (in body weight units), participant IDs, jump height (m) and peak power (W·kg⁻¹). These data support the following paper submitted to Sports Biomechanics: "Reconstructing ground reaction force curves from a single accelerometer using functional principal components."