MultiplEYE Preprocessing

Tip

Version v2026.07.01 (July 2026) marks the first stable release. From this version onward, the output format and core API are considered stable. Future updates will focus on maintenance, bug fixes, and feature additions without breaking changes.

Important

This repository is stable and actively maintained. Please:

• Keep the repository up-to-date to receive the latest changes, fixes, and improvements
• Report issues if you encounter unexpected behavior
• Refer to the documentation for the latest information on how to use the pipeline
• Check the
• troubleshooting guide if you encounter any issues

This repository contains the preprocessing pipeline for eye-tracking data and psychometric test scoring from the MultiplEYE project.

If you are running the pipeline and encounter any issues, please check the troubleshooting guide.

Note

This repository processes data recorded with MultiplEYE-psychometric-tests.

What You Can Do With This Repository

1. Run the Preprocessing Pipeline

Process raw eye-tracking data (EyeLink .edf files) and score psychometric tests in one workflow:

# Configure your settings in multipleye_settings_preprocessing.yaml
run_preprocessing

This pipeline handles:

• Converting .edf to .asc format
• Parsing and validating eye-tracking data
• Applying filters and detecting events
• Generating preprocessed output files
• Scoring psychometric tests

Tip

A step-by-step notebook is available in preprocessing.ipynb to walk through the pipeline in detail.

Important

Psychometric tests require data to be structured correctly. See the Psychometric Tests documentation for details on the expected data format.

2. Score Psychometric Tests (Standalone)

If you only need to re-score psychometric tests without re-running the full pipeline, use the standalone command:

preprocess_psychometric_tests

This uses the same data already prepared by run_preprocessing and produces the same output.

---

Installation

For full installation instructions, see the Getting Started guide.

Quick setup:

git clone https://github.com/MultiplEYE-COST/multipleye-preprocessing.git
cd multipleye-preprocessing/
uv sync
source .venv/bin/activate  # Unix/Mac
# or
.venv\Scripts\activate  # Windows

---

Documentation Overview

Topic	Description
Getting Started	Installation, requirements, and running the pipeline
Preprocessing	Detailed preprocessing pipeline documentation
Reading Measures	Reading measures from preprocessed eye-tracking data
Psychometric Tests	Test descriptions and scoring details
Configuration	Configuration file options
Technical Architecture	Code structure and design

---

Quick Start

1. Update settings in multipleye_settings_preprocessing.yaml
2. Run the full pipeline (eye-tracking + psychometric tests): run_preprocessing
3. Re-score psychometric tests standalone (optional): preprocess_psychometric_tests

Caution

EyeLink-specific: You must install the EyeLink Developers Kit to convert .edf files. See the installation guide for details.