#275 Wire RQ1/RQ2 benchmark matrix

[jathavaan]

Rewrites benchmarks.yml to the 52 active experiments from Table 4.2.1 (RQ1: 4 query types x 7 size variants = 28; RQ2: 6 single-node + 18 Sedona = 24). RQ1 pairs three engines at small and two at medium/large via related_script_ids; Sedona variants run unpaired.

Adds _strip_dataset_size_suffix in benchmark_runner.py so 52 unique yml ids (each carrying a -small/-medium/-large tail) dispatch to the shared base entrypoint while dataset_size flows in via --dataset-size.

Collapses docker-compose.yml to 20 active services (one per (query_type, engine) pair) and trims both CI build matrices to match. Out-of-scope benchmarks stay in tree but are commented out at the bottom of benchmarks.yml, docker-compose.yml, and the workflow matrices so re-enabling them is a single search-and-uncomment.

Adds a Test matrix table to the README documenting which experiments launch as a parallel pair group and which run alone.

[Copilot]

Pull request overview

Reorganizes the benchmark configuration to align with Table 4.2.1 (52 active experiments: 28 for RQ1, 24 for RQ2) by introducing per-size experiment ids while keeping a single entrypoint per (query_type, engine) pair. The orchestrator already forwards --dataset-size separately, so a small suffix-stripping helper in benchmark_runner.py keeps the dispatch table compact. Out-of-scope items are commented out (not deleted) in all four affected files so re-enabling is uniform.

Changes:

• Rewrites benchmarks.yml to 52 size-suffixed experiments with correctly scoped related_script_ids (3-way at small, 2-way at medium/large, Sedona unpaired); preserves out-of-scope entries as comments.
• Adds _strip_dataset_size_suffix in benchmark_runner.py so size-suffixed ids dispatch to the existing base cases; the match table now correctly covers every active base id (verified all 12 RQ1 + 8 RQ2 base ids exist as cases).
• Trims docker-compose.yml and the two GitHub Actions matrices to the 20 active services, and extends the README with a Test matrix section documenting pair groups (12 RQ1 + 21 RQ2 = 33, consistent with 52 experiments).

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
benchmarks.yml	Replaces flat experiment list with 52 size-suffixed entries; mutual related_script_ids are symmetric at each size; commented out-of-scope block.
benchmark_runner.py	Adds _strip_dataset_size_suffix and wires it into the dispatch match; iterates DatasetSize enum so small/medium/large suffixes are stripped before lookup.
docker-compose.yml	Reduces to 20 active services (one per base id) plus commented out-of-scope mirror.
.github/workflows/push-containers-to-acr.yml	Trims build matrix to the 20 active images; commented mirror at the bottom.
.github/workflows/pull-request-tests.yml	Mirrors the trimmed service matrix for PR tests.
README.md	Adds TOC entry and "Test matrix" subsection documenting RQ1/RQ2 pair groupings and Shapefile-only-at-small policy.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.