BTU Graph data

Versioned, machine-readable snapshots of the BTU Graph energy company knowledge graph. The dataset connects energy companies, sectors, capabilities, markets, regions, workflows, software products, market-data requirements, and operational decisions.

The canonical dataset page is btugraph.com/data. Entity URLs in the exports point to the corresponding canonical BTU Graph profiles.

Current snapshot

• Edition: 2026-09-04
• Schema version: 4
• Companies: 66
• Graph nodes: 286
• Graph relationships: 1085
• Access: no authentication required
• Formats: JSON, CSV, and OpenAPI 3.1

File	Description	Canonical live source
data/graph.json	Full typed graph, provenance, data dictionary, and evidence links	/graph.json
data/companies.csv	Flat company directory for analysis and spreadsheets	/companies.csv
data/openapi.json	OpenAPI description of all public BTU Graph data endpoints	/openapi.json
data/metadata.json	Snapshot identity, counts, source endpoints, and SHA-256 digests	Generated from the files above
SHA256SUMS	Integrity manifest for the snapshot	Generated by the sync script

Quick start

Download the current repository snapshot:

curl -O https://raw.githubusercontent.com/OilpriceAPI/btugraph-data/main/data/graph.json

List companies associated with commodity trading and origination:

jq -r '
  .edges[]
  | select(.relationship == "has_capability" and .to == "capability:commodity-trading-origination")
  | .from
' data/graph.json

Run the dependency-free examples:

npm run example:javascript
npm run example:python

See docs/DATA_DICTIONARY.md for the model and docs/QUERIES.md for more examples.

Provenance and editorial method

BTU Graph compiles public evidence into source-reviewed entity records and editorial relationship classifications. Company nodes include the public sources used during review. A lastReviewed date means the cited evidence was checked on that date; it is not a real-time status guarantee.

• Methodology
• Canonical data page
• Company directory
• Corrections

Releases and updates

Git tags and GitHub releases identify immutable editions. A scheduled workflow checks the canonical exports weekly and opens a pull request only when the source data changes. Every change must pass the same integrity checks as a local sync.

To refresh a working copy from the canonical endpoints:

npm run sync

License and attribution

BTU Graph-authored directory metadata, descriptions, and editorial relationship classifications are licensed under Creative Commons Attribution 4.0 International. Attribute the dataset as:

BTU Graph, Oil Price API — https://btugraph.com/data/

Company names, trademarks, logos, and third-party source materials remain the property of their respective owners and are not included in that license grant. Evidence URLs identify sources; this repository does not redistribute those source materials. See NOTICE.md for the complete scope statement.

Corrections and contributions

Submit factual dataset corrections through the canonical correction process with a public supporting source. Repository improvements to schemas, documentation, examples, and validation are welcome; see CONTRIBUTING.md.