docs(gfql): benchmarks vs Spark GraphFrames + LadybugDB (single-node, honest)#1668

Merged

lmeyerov merged 12 commits into

feat/gfql-graphframes-bench

Jul 9, 2026

Commits on Jul 9, 2026

docs(gfql): benchmark GFQL vs Spark GraphFrames (L4, single-node, honest)

lmeyerov
and
claude
committed
docs(gfql): mark illustrative benchmark_graphframes snippets doc-test:skip

lmeyerov
and
claude
committed
docs(gfql): correct Friendster framing — eager-load harness limit, not an engine ceiling

lmeyerov
and
claude
committed
bench(gfql): LadybugDB (Kuzu fork) harness + engines.rst row (WIP, tiny-validated)

lmeyerov
and
claude
committed
docs(gfql): Ladybug row — GFQL has streaming, don't concede out-of-core

lmeyerov
and
claude
committed
bench(gfql): fair GFQL-vs-Ladybug Cypher benchmark (their ops, row-pipeline path)

lmeyerov
and
claude
committed
bench(gfql): Ladybug harness — build native-per-engine (fair, no per-call conversion)

lmeyerov
and
claude
committed
docs(gfql): LadybugDB row — real head-to-head numbers (native-per-engine, honest)

lmeyerov
and
claude
committed
docs+bench(gfql): review fixes — cross-machine disclosure, honest labels, ranges

lmeyerov
and
claude
committed
docs(gfql): pdflatex rejects Greek ρ (U+03C1) — spell out 'rho'

lmeyerov
and
claude
committed
bench(gfql): fail graphframes parity mismatches
lmeyerov
committed
docs(gfql): harden graphframes benchmark wording
lmeyerov
committed