EEGUnity documentation

Contents:

• API Reference
  • eegunity package
    • Core API
    • eegunity.unifieddataset module
      • UnifiedDataset
    • Subpackages
      • eegunity.modules package
      • eegunity.resources package
      • eegunity.utils package
    • Top-level exports
• Tutorials
  • How to Support Multi-Modal Data in EEGUnity
    • Recommended Workflow
    • About misc: Label Channels
    • Custom Processing for Multi-Modal Workloads
    • Notes
  • Building File Hash and File Size Metadata
    • Metadata Columns
    • Step 1: Compute Integrity Metadata
    • Step 2: Find Potential Duplicates
    • Step 3: Find Missing/Unreadable Files
    • Step 4: Save the Enriched Locator
    • Notes
  • Formatting Channel Names and Inspecting Metadata
    • Requirements
    • Step 1: Load the Dataset
    • Step 2: Save the Initial Locator
    • Step 3: Keep Completed Records
    • Step 4: Format Channel Names
    • Step 5: Manual Correction (Optional)
    • Step 6: Reload from Locator
  • Creating a Standardized FIF Dataset with EEGUnity
    • Prerequisites
    • Recommended Project Layout
    • Step 1: Build Initial Locator
    • Step 2: Define Dataset-Specific Conversion Logic
    • Step 3: Validate Converted Dataset
    • Notes
  • Parsing Non-standard Data Formats in EEGUnity
    • Supported Formats
    • Step 1: Build Locator with Parser Extensions Enabled
    • Step 2: Inspect Specific File Types
    • Step 3: Load a Non-standard Row with get_data_row
    • Step 4: Batch Validate Readability
    • Notes
  • Working Across Multiple Computers with EEGUnity
    • Scenario
    • Step 1: Researcher A Generates Locator and Integrity Metadata
    • Step 2: Researcher B Replaces Local Paths
    • Step 3: Researcher B Recomputes and Verifies Consistency
    • Notes
  • EEG Dataset Processing and h5Dataset Export
    • Workflow A: Step-by-Step Pipeline
    • Workflow B: Quick Custom Pipeline with batch_process
    • Notes
  • Reading EEGUnity HDF5 Files
    • HDF5 Layout Used by EEGUnity
    • Example Script
    • Optional: Load Only Metadata
    • Notes
  • EEGUnity Kernel Tutorial: Rich Metadata, misc, stim, and Annotations
    • 1. Design Principle
    • 2. What a Kernel Can Do
      • Standard kernel interface
    • 3. Annotation vs misc vs stim
    • 4. Example Kernel with Multiple misc and stim Channels
    • 5. Binding and Running
    • 6. Channel Type Compatibility
    • 7. Extended Interface: Handling Unavailable Files
      • Call sequence for Unavailable files
      • Example: headerless CSV dataset
      • Backward compatibility
    • 8. Recommended Practice
  • How to Speed Up EEGUnity with Built-in Parallelism
    • 1. Introduction
    • 2. Basic Configuration
    • 3. Execution Modes in batch_process
    • 4. Where Parallelism Is Commonly Applied
      • 4.1 Parser Stage
      • 4.2 Batch Stage
    • 5. Choosing Worker Count
    • 6. Practical Tips
    • 7. Summary

Indices and tables

• Index

• Module Index

• Search Page