{% extends "base.html" %} {% block title %}Datasets{% endblock %} {% block content %}

Datasets

Versions and supersession chains. A dataset referenced by any completed experiment is immutable: a fix creates a new version rather than editing the old one, so results stay attached to the data they were computed on.

{% if datasets %}
{% for d in datasets %} {% endfor %}
DatasetVersionStatusOrigin StorageRowsCreatedChain
{{ d.title }} {{ d.description }} {% if d.defect %}{{ d.defect }}{% endif %} {{ d.version }} {{ d.status }} {% if d.generated %}generated {{ d.generator }} {% else %}source{% endif %} {{ d.storage }} {{ d.row_count | num }} {{ d.created or '—' }} {% if d.supersedes_link %}replaces {{ d.supersedes_link.label }}{% endif %} {% if d.superseded_by_link %}replaced by {{ d.superseded_by_link.label }}{% endif %} {% if not d.supersedes_link and not d.superseded_by_link %}—{% endif %}

A generated dataset carries storage: none and is gitignored: if it can be regenerated it does not belong in the repository, and the generator plus the command in its Provenance section is the record. Source data that cannot be regenerated is committed via git LFS so the repo has no external dependency.

{% else %}

No datasets referenced yet. They appear when an experiment points at data.

{% endif %} {% endblock %}