How-To Guides#

Estimated reading time:3 minutes

Task-focused recipes for specific EEGDash workflows. Each guide assumes you already know the basics and want a direct answer to a single question – “how do I work offline?”, “how do I parallelize feature extraction?”, “how do I run preprocessing on SLURM?”. Difficulty 1-2; assumes the Start Here trio.

How-to guides sit in the recipe quadrant of the Diataxis framework: not a curated learning path (those are the tutorials), not a deep explanation (that is Concepts), not a complete API reference. They are the answers to the operational questions that come up once you are running EEGDash for real work. Cross-link with the HPC track when relevant.

What you will learn:

How to download a dataset locally and pin it in the cache so reruns do not refetch.
How to parallelize feature extraction across CPU cores using joblib and EEGDash’s batch helpers.
How to run preprocessing as a SLURM array job on a shared cluster (paired with the HPC tutorials).
How to use the HPC cache layout so two jobs on the same cluster share preprocessed data.
How to work fully offline: cache management, manifest export, and reloading without network access.

Each how-to is a single self-contained script or markdown file.

Download a real EEG subset and verify it can be reopened

Measure parallel feature extraction on recorded EEG

Stage a real EEG recording onto job-local storage

Read recorded EEG without a metadata or download connection