# EEGDash

> EEGDash is an open catalog and Python library for finding, loading, and
> preprocessing publicly available EEG, MEG, and iEEG datasets in BIDS
> format. It aggregates recordings from OpenNeuro, NEMAR, Zenodo,
> Figshare, SciDB, OSF, DataRN, and EEGManyLabs into a single searchable
> catalog with a uniform `EEGDashDataset` Python interface compatible
> with MNE-Python and braindecode.

The project is maintained by the BrainIAK / EEGDash contributors and is
released under an open-source license. Dataset licenses are inherited
from each upstream source and must be checked independently.

Automated clients: prefer the structured resources listed under
"Machine-readable" before parsing HTML pages.

## Getting started

- [Project homepage](https://eegdash.org/index.html): what EEGDash is and why.
- [Install with pip](https://eegdash.org/install/install_pip.html): `pip install eegdash`.
- [Install from source](https://eegdash.org/install/install_source.html): developer setup.
- [User guide](https://eegdash.org/user_guide.html): narrative walkthrough of the core workflow.

## Catalog

- [Dataset summary](https://eegdash.org/dataset_summary.html): interactive catalog of every dataset with counts, modalities, tasks, and links to upstream archives.

## Python API

- [API overview](https://eegdash.org/api/api.html): top-level entry points.
- [Core API reference](https://eegdash.org/api/api_core.html): `eegdash.api`, schemas, downloader, HTTP client.
- [Dataset class reference](https://eegdash.org/api/dataset/eegdash.EEGDashDataset.html): filters, lazy loading, BIDS metadata.
- [Feature-extraction overview](https://eegdash.org/api/features_overview.html): spectral, connectivity, bivariate, and complexity features.
- [Features API reference](https://eegdash.org/api/api_features.html): `eegdash.features` module listings.

## Tutorials and examples

- [All tutorials index](https://eegdash.org/generated/auto_examples/index.html)
- [Minimal tutorial](https://eegdash.org/generated/auto_examples/core/tutorial_minimal.html): load a dataset and train a tiny model end-to-end.
- [Eyes-open / eyes-closed tutorial](https://eegdash.org/generated/auto_examples/core/tutorial_eoec.html): classic EEG classification.
- [P300 transfer learning](https://eegdash.org/generated/auto_examples/core/p300_transfer_learning.html): cross-subject transfer on the P300 paradigm.
- [Feature extraction on EOEC](https://eegdash.org/generated/auto_examples/core/tutorial_feature_extractor_open_close_eye.html): using the feature API on resting-state data.
- [Age prediction tutorial](https://eegdash.org/generated/auto_examples/tutorials/noplot_tutorial_age_prediction.html): regression from raw EEG.
- [p-factor regression](https://eegdash.org/generated/auto_examples/tutorials/noplot_tutorial_pfactor_regression.html): clinical outcome regression.
- [Auditory oddball](https://eegdash.org/generated/auto_examples/tutorials/noplot_tutorial_audi_oddball.html): event-related paradigm.

## Project info

- [Developer notes](https://eegdash.org/developer_notes.html): contribution, build, and release notes.
- [GitHub repository](https://github.com/eegdash/EEGDash): source code and issue tracker.

## Machine-readable

- [Agent Skills manifest](https://eegdash.org/.well-known/agent-skills/index.json): structured skills (find datasets, get metadata, load BIDS records, count records, list features).
- [API catalog (RFC 9727)](https://eegdash.org/.well-known/api-catalog): linkset pointing at the public EEGDash HTTP API.
- [Full markdown corpus](https://eegdash.org/llms-full.txt): concatenation of every rendered markdown page (larger, for full-corpus retrieval).
- [OpenAPI specification](https://data.eegdash.org/openapi.json): full OpenAPI 3.1 spec for the `data.eegdash.org` catalog API.
- [Swagger UI](https://data.eegdash.org/docs) and [ReDoc](https://data.eegdash.org/redoc): human-readable API documentation.
- [Sitemap](https://eegdash.org/sitemap.xml): every indexable page on this site.
- [robots.txt](https://eegdash.org/robots.txt): crawl rules and Content Signals (`search=yes, ai-input=yes, ai-train=no`).

## Optional

- [BIDS specification](https://bids-specification.readthedocs.io/): the data format EEGDash speaks natively.
- [MNE-Python](https://mne.tools/): the numerical backbone used by `EEGDashDataset`.
- [braindecode](https://braindecode.org/): downstream deep-learning library compatible with EEGDash outputs.

## API reference pages (per-module)


## Dataset pages (N=6)

Every EEGDash dataset has a dedicated Sphinx page with its BIDS metadata, upstream citation, and load-in-Python snippet. The full interactive catalog lives at <https://eegdash.org/dataset_summary.html>; the list below is what sitemap / agent crawlers should enumerate.

- [base](https://eegdash.org/api/dataset/eegdash.dataset.base.html)
- [bids_dataset](https://eegdash.org/api/dataset/eegdash.dataset.bids_dataset.html)
- [dataset](https://eegdash.org/api/dataset/eegdash.dataset.dataset.html)
- [exceptions](https://eegdash.org/api/dataset/eegdash.dataset.exceptions.html)
- [io](https://eegdash.org/api/dataset/eegdash.dataset.io.html)
- [registry](https://eegdash.org/api/dataset/eegdash.dataset.registry.html)

---