Interoperability, Eye-Tracking-BIDS, and storage

Eye-Tracking-BIDS

export_eye_bids(
  dataset,
  "bids-eye-study",
  task = "reasoning",
  screen_distance_m = 0.60,
  screen_size_m = c(0.53, 0.30),
  overwrite = TRUE
)
roundtrip <- import_eye_bids("bids-eye-study")
validate_eye_dataset(roundtrip)

The exporter writes one physiological recording per eye, with timestamp, x_coordinate, and y_coordinate first, optional pupil_size, no TSV header, and required column metadata in the JSON sidecar.

Parquet and Arrow

handle <- write_eye_storage(dataset, "storage", format = "parquet", overwrite = TRUE)
handle
restored <- collect_eye_storage(handle)

Package ecosystems

x <- as_eyeprocess_eyetools(external_object, mapping = declared_mapping)
proc <- as_procdata_sequence(x)
seqs <- as_traminer_sequence(x, create_object = TRUE)
hmm_data <- as_seqhmm_data(x)