Composite Datasets

Compose a Hierarchical Dataset

Merge two subsystem datasets into a parent composite. Items and relations from both sources are combined. An optional inter-system coupling layer is added to connect nodes across the two subsystems.

Examples

human_bodybrain + heart + immune_system · inter: autonomic_coupling

mars_missionlaunch_system + surface_ops · inter: resource_dependency

software_projectfrontend + backend · inter: api_coupling

citytransport + power_grid · inter: spatial_co_location

Subsystem A

Subsystem B

Inter-system coupling layer name(the new layer type added between A and B)

Inter-system edges — optional JSON array of { from, to, weight? } objects

from = node ID from Subsystem A · to = node ID from Subsystem B. Leave blank to compose without cross-system edges.

Prefix for A node IDs— only needed if IDs clash

Prefix for B node IDs

Composite name *

Domain *

Version

Description

How composite datasets work

1. Select two subsystem datasetsEach child dataset is a full multilayer graph (items + relations). They can be different domains or subsystems of the same parent system.
2. Define the inter-system layerName the coupling type (e.g. resource_dependency, api_coupling, spatial_co_location). Then add specific edges between nodes across the two subsystems.
3. ComposeIRDME merges items and relations from both sources, appends the inter-system coupling layer, and stores the result as a new composite dataset with full provenance (childDatasetIds, version, compositionLayer).
4. Analyze the compositeRun any IRDME analysis on the composite dataset — structural audit, dataset comparison, atlas, law. The inter-system layer exposes hub shadows that neither subsystem reveals alone.

Composite datasets are versioned (semver). Child dataset IDs are stored for full provenance. Run Dataset Integrity Comparison to verify structural agreement between the composite and individual sources.