surface↓ranking↓pagerank
Rankingscalar · returns json

PAGERANK

PageRank centrality on an ad-hoc edge list (NetworkX)

Per-row — runs once for each row.

rankingdeterministicjson

Arguments

nametypedescription
edgesJSON—
damping(optional)DOUBLE—

About

PageRank on an ad-hoc edge table, as a SQL table function. DuckDB has zero native graph operators. This cascade fills the gap by taking a JSON array of edges `[[src, dst, weight?], ...]` and running NetworkX's PageRank algorithm on it in-process, returning one row per node with its score. Uses the `returns_columns` TVF pattern — the registry auto-generates a `pagerank_rows(...)` macro callable with FROM / LATERAL. Usage: -- Compute PageRank from an edge table SELECT * FROM pagerank_rows( (SELECT JSON_GROUP_ARRAY(JSON_ARRAY(src, dst, weight)) FROM edges) ) ORDER BY score DESC LIMIT 20; Edge weights are optional — missing weights default to 1.0. The algorithm treats the graph as directed; symmetrize your edges first if you want an undirected PageRank.

Examples

PageRank on a tiny 4-node directed graph (top node)

SELECT
  node,
  rank
FROM
  pagerank_rows (
    JSON(
      '[["A","B"],["B","C"],["C","A"],["A","C"],["D","A"]]'
    )
  )
ORDER BY
  score DESC
LIMIT
  1;

Nearby rabbit holes

same domain
scalar

ABOUT

Returns 0.0-1.0 relevance score for text vs criterion (cross-encoder)

rankingrerankerspecialist-zootext
scalar

ABOUT_LLM

LLM-backed 0.0-1.0 relevance score (escape hatch for ABOUT/RELEVANCE TO)

rankingllmllm-escape-hatchtext
aggregate

BEST

Pick the single best value from a group by a plain-English quality criterion

rankingllmtext
aggregate

OUTLIERS

Find unusual or atypical items via embeddings (+ optional criteria)

rankingrerankerspecialist-zoojson
aggregate

OUTLIERS_LLM

LLM-backed outlier detection (escape hatch for OUTLIERS)

rankingllmllm-escape-hatchjson
aggregate

RANK_SEMANTIC

Rank a group of items by a subjective multi-factor criterion

rankingllmtext
Climb back to The Looking Glass