Documentation Overview
Task-first documentation for installing EyeSort 1.0, labeling synchronized fixation events, and validating the outputs.
Last updated: 2026-10-07
Start with the outcome
EyeSort 1.0 turns synchronized fixation events in continuous EEGLAB datasets into reproducible, behavior-contingent event markers. It maps horizontal fixation positions to user-defined text regions, derives reading context, and labels only the fixation events that match a reviewed label queue.
EyeSort belongs after synchronization and eye-event detection, and before ERP/FRP averaging, BINLISTER, time-frequency analysis, or deconvolution. It does not synchronize streams, preprocess EEG, detect or repair fixations, blinks, or missing eye data, or separate temporally overlapping neural activity.
Current scope
EyeSort 1.0 is designed for synchronized, continuous EEGLAB .set data from single-line reading studies. It reconstructs user-defined text regions from one horizontal sentence offset, literal character counts, and uniform pixels per character. Mapping uses fixation X position only.
The verified design envelope is a fixed-width font, consistent display geometry, numeric condition/item triggers, and the same named region structure across relevant trials. The labeler selects those user-defined regions; it does not expose every calculated word as an independently selectable region.
Saccades provide incoming/outgoing direction context, but EyeSort labels fixation events only. If your data still need synchronization, fixation detection, blink/missing-data repair, or EEG preprocessing, complete those upstream steps first.
Choose a path
- New installation: check requirements, then install or upgrade.
- First evaluation: follow the sample-data quickstart.
- Own data: prepare the EEG events and IA file, then follow the single-dataset guide.
- Many participants: use the batch workflow.
- Existing outputs: consult the event fields, output files, or diagnostics.
Verified 1.0 menu sequence
- 1. Load EEG Dataset(s) — choose one or more
.setfiles and a required output directory. - 2. Setup Interest Areas — configure text geometry; optionally save
*_eyesort_ia.set. - Optionally Inspect Parsed Regions and 3. Import IA Columns to Events.
- 4. Eye-Tracking Event Labeling — queue labels and apply them. EyeSort auto-saves
*_processed.set/.fdt. - Optionally generate
eyesort_bins.txt, convert the displayed event-marker format, and export a processing script.
There is no “Save Labeled Dataset” menu command in EyeSort 1.0.