Explore collection lineage

Explore collection lineage

Datasets (DataObjects) in a collection can be linked by two types of relationship:

• Parent / child — a dataset is structurally contained within another (e.g. a test run within a campaign).
• Predecessor / successor — a dataset was produced from another in a processing step (e.g. a processed result derived from raw data).

Lineage graph

1. Open a Collection.
2. Scroll to the Dataset Lineage panel and expand it.
3. The graph renders automatically once the datasets load.

Reading the graph:

Interacting:

• Scroll to zoom in/out — the graph adapts to the zoom level (process-type bubbles when zoomed far out, individual datasets when zoomed in). See Navigate a large lineage graph for the full zoom-mode / filter / minimap walkthrough.
• Drag the canvas to pan around.
• Hover a node to see the dataset name, status, and description.
• Click a node to open its detail page.
• Use the filter pills above the graph to scope by status or process type without reloading.

Setting relationships

Relationships are set when creating or editing a dataset:

1. Open or create a dataset.
2. In the Edit dialog (pencil icon), scroll to the Relationships section.
3. Set the Parent or Predecessors fields.

Changes appear in the lineage graph after the panel refreshes.

Jumping between sibling datasets

The left sidebar on every collection page is also the sibling navigator: click any dataset in the tree to jump straight to it, and use the chevron (›) next to a dataset to expand its children (for example, an anomaly’s investigation sub-tree). Hover the small information icon next to the Contents header for a reminder of this shortcut.