Skip to contents

Read GaitRec Dataset

Source: R/io-gaitrec.R
readGaitRec.Rd

Reads ground reaction force (GRF) data or spatiotemporal parameters from the GaitRec dataset format (Horsak et al., 2020). GaitRec is a large-scale clinical gait dataset containing GRF waveforms and gait parameters from patients with various lower-limb orthopaedic conditions and healthy controls.

Usage

readGaitRec(
  path,
  type = c("grf", "parameters"),
  sr = 1000,
  sep = ",",
  subject_id = NULL
)

Arguments

path

Path to a GaitRec CSV file or a directory containing GaitRec CSV files. When a directory is given, all CSV files matching the specified type are read and combined.

type

Type of data to read: "grf" for ground reaction force waveforms or "parameters" for spatiotemporal parameters. Default "grf".

sr

Sampling rate in Hz. Default 1000 for GaitRec GRF data. For time-normalised waveforms (101 data points representing 0–100\ gait cycle), this value is used as a nominal sampling rate in the resulting PhysioExperiment object.

sep

Column separator in the CSV file. Default ",". Set to "\t" for tab-separated files.

subject_id

Optional subject identifier string to store in the returned object's metadata. When path is a directory this is ignored and identifiers are extracted from file names.

Value

For type = "grf": a PhysioExperiment object with a "raw" assay containing the GRF data matrix (time points x channels). The colData records channel labels, force component ("Fx", "Fy", "Fz"), plate number, and unit ("N"). Metadata includes source_file, format, and subject_id.

For type = "parameters": a data.frame with spatiotemporal gait parameters extracted from the CSV file.

Details

GaitRec GRF file format

GaitRec GRF CSV files typically contain columns for time and force components from two force plates:

  • A time column (e.g., "time", "Time", "t")

  • Force columns following naming patterns like "Fx1", "Fy1", "Fz1", "Fx2", "Fy2", "Fz2" (plate number suffix) or "FP1_Fx", "FP1_Fy", "FP1_Fz", "FP2_Fx", etc. (plate prefix)

The function auto-detects force column naming patterns and extracts the component (x/y/z) and plate number. Columns that do not match recognised force patterns are stored in metadata(pe)$extra_columns as a data.frame.

Time-normalised waveforms (101 data points, 0–100\ also supported. The parser detects these by row count and stores a percent_gait_cycle vector in the metadata.

References

Horsak B, Slijepcevic D, Raberger A-M, Schwab C, Worisch M, Zeppelzauer M (2020). "GaitRec, a large-scale ground reaction force dataset of healthy and impaired gait." Scientific Data, 7, 143. doi:10.1038/s41597-020-0481-z

See also

readMoCapCSV() for generic CSV motion capture data, filterGRF() for low-pass filtering GRF signals, analyzeForcePlate() for force plate analysis.

Examples

if (FALSE) { # \dontrun{
# Read a single GRF file
pe <- readGaitRec("path/to/gaitrec_grf.csv")

# Read spatiotemporal parameters
params <- readGaitRec("path/to/gaitrec_params.csv", type = "parameters")

# Read with tab separator
pe <- readGaitRec("path/to/gaitrec.tsv", sep = "\t")
} # }