Add benchmark results and update scripts for OLAP and vector operations

- Introduced new benchmark result files for graph OLAP, vector index build, and vector search across all dataset sizes.
- Updated scripts for running OLTP and OLAP matrix tests to reflect changes in database identifiers, specifically replacing "arcadedb" with "arcadedb_sql".
- Enhanced version management in summary scripts to accommodate new database identifiers.
- Adjusted dataset configurations and transaction parameters for improved testing accuracy and performance.

Author: tae898

bindings/python/examples/07_stackoverflow_tables_oltp.py

@@ -2039,7 +2039,7 @@ def run_in_docker(args):
         filtered_args.append(arg)
 
     arcadedb_wheel_mount_path = None
-    if args.db == "arcadedb":
+    if args.db == "arcadedb_sql":
         wheel_candidates = sorted(
             (repo_root / "bindings/python/dist").glob("*embed*.whl")
         )
@@ -2147,9 +2147,9 @@ def main():
     )
     parser.add_argument(
         "--db",
-        choices=["arcadedb", "sqlite", "duckdb", "postgresql"],
-        default="arcadedb",
-        help="Database to test (default: arcadedb)",
+        choices=["arcadedb_sql", "sqlite", "duckdb", "postgresql"],
+        default="arcadedb_sql",
+        help="Database to test (default: arcadedb_sql)",
     )
     parser.add_argument(
         "--threads", type=int, default=4, help="Number of worker threads (default: 4)"
@@ -2208,7 +2208,7 @@ def main():
 
     heap_size = (
         resolve_arcadedb_heap_size(args.mem_limit, args.jvm_heap_fraction)
-        if args.db == "arcadedb"
+        if args.db == "arcadedb_sql"
         else None
     )
     args.heap_size_effective = heap_size
@@ -2234,12 +2234,12 @@ def main():
         print(f"SQLite profile: {args.sqlite_profile}")
     print(f"Threads: {args.threads}")
     print(f"Transactions: {args.transactions:,}")
-    if args.db == "arcadedb":
+    if args.db == "arcadedb_sql":
         print(f"JVM heap size: {heap_size}")
     print(f"DB path: {db_path}")
     print()
 
-    if args.db == "arcadedb":
+    if args.db == "arcadedb_sql":
         summary = run_oltp_arcadedb(
             db_path=db_path,
             data_dir=data_dir,

bindings/python/examples/08_stackoverflow_tables_olap.py

@@ -1349,7 +1349,7 @@ def run_in_docker(args):
         filtered_args.append(arg)
 
     arcadedb_wheel_mount_path = None
-    if args.db == "arcadedb":
+    if args.db == "arcadedb_sql":
         wheel_candidates = sorted(
             (repo_root / "bindings/python/dist").glob("*embed*.whl")
         )
@@ -1995,9 +1995,9 @@ def main():
     )
     parser.add_argument(
         "--db",
-        choices=["arcadedb", "sqlite", "duckdb", "postgresql"],
-        default="arcadedb",
-        help="Database to test (default: arcadedb)",
+        choices=["arcadedb_sql", "sqlite", "duckdb", "postgresql"],
+        default="arcadedb_sql",
+        help="Database to test (default: arcadedb_sql)",
     )
     parser.add_argument(
         "--batch-size",
@@ -2084,12 +2084,12 @@ def main():
             args.mem_limit,
             args.jvm_heap_fraction,
         )
-        if args.db == "arcadedb"
+        if args.db == "arcadedb_sql"
         else args.mem_limit
     )
     args.heap_size_effective = heap_size
     jvm_kwargs = {"heap_size": heap_size}
-    if args.db == "arcadedb":
+    if args.db == "arcadedb_sql":
         jvm_kwargs["jvm_args"] = (
             "-Darcadedb.queryMaxHeapElementsAllowedPerOp="
             f"{args.arcadedb_query_max_heap_elements}"
@@ -2112,7 +2112,7 @@ def main():
     print(f"Query runs: {args.query_runs}")
     print(f"Query order: {args.query_order}")
     print(f"Seed: {args.seed}")
-    if args.db == "arcadedb":
+    if args.db == "arcadedb_sql":
         print(f"JVM heap size: {heap_size}")
         print(
             "ArcadeDB query max heap elements/op: "
@@ -2123,7 +2123,7 @@ def main():
 
     start_time = time.perf_counter()
 
-    if args.db == "arcadedb":
+    if args.db == "arcadedb_sql":
         stop_event, rss_state, rss_thread = start_rss_sampler()
         summary = run_olap_arcadedb(
             db_path=db_path,
@@ -2180,7 +2180,7 @@ def main():
         total_time = time.perf_counter() - start_time
     else:
         raise NotImplementedError(
-            "Only arcadedb, sqlite, duckdb, and postgresql are supported for now"
+            "Only arcadedb_sql, sqlite, duckdb, and postgresql are supported for now"
         )
 
     summary["total_time_s"] = total_time

bindings/python/examples/09_stackoverflow_graph_oltp.py

@@ -548,13 +548,9 @@ def verify_single_thread_series(
     verify_dir = db_path
     verify_dir.mkdir(parents=True, exist_ok=True)
     dataset_tag = args.dataset.replace("-", "_")
-    arcadedb_oltp_language = (
-        args.arcadedb_oltp_language if db_label == "arcadedb" else None
-    )
-    language_suffix = f"_{arcadedb_oltp_language}" if arcadedb_oltp_language else ""
     baseline_path = (
         verify_dir
-        / f"verify_09_graph_oltp_{db_label}{language_suffix}_{dataset_tag}_tx{args.transactions}_seed{args.seed}.json"
+        / f"verify_09_graph_oltp_{db_label}_{dataset_tag}_tx{args.transactions}_seed{args.seed}.json"
     )
 
     current_payload = {
@@ -563,7 +559,6 @@ def verify_single_thread_series(
         "transactions": int(args.transactions),
         "seed": int(args.seed),
         "threads": int(args.threads),
-        "arcadedb_oltp_language": arcadedb_oltp_language,
         "node_count": int(summary.get("node_count") or 0),
         "edge_count": int(summary.get("edge_count") or 0),
         "node_counts_by_type": normalize_count_dict(summary.get("node_counts_by_type")),
@@ -578,7 +573,6 @@ def write_baseline() -> None:
                     "%Y-%m-%dT%H:%M:%SZ"
                 ),
                 "source_db": db_label,
-                "source_oltp_language": arcadedb_oltp_language,
             },
             "baseline": current_payload,
         }
@@ -2435,7 +2429,7 @@ def run_graph_oltp_arcadedb(
     next_answer_id = load_info["max_ids"]["answer"] + 1
     next_badge_id = load_info["max_ids"]["badge"] + 1
     next_comment_id = load_info["max_ids"]["comment"] + 1
-    arcadedb_oltp_language = (oltp_language or "cypher").strip().lower()
+    arcadedb_mode = (oltp_language or "cypher").strip().lower()
 
     def worker(ops: List[str], worker_id: int) -> Dict[str, List[float]]:
         rng = random.Random(seed + worker_id)
@@ -2449,7 +2443,7 @@ def worker(ops: List[str], worker_id: int) -> Dict[str, List[float]]:
         for op in ops:
             start_time = time.perf_counter()
 
-            if arcadedb_oltp_language == "sql":
+            if arcadedb_mode == "sql":
                 try:
                     if op == "read":
                         read_kind = rng.choice(READ_TARGET_KINDS)
@@ -3640,7 +3634,6 @@ def do_delete():
         "edge_count": edge_count,
         "node_counts_by_type": node_counts_by_type,
         "edge_counts_by_type": edge_counts_by_type,
-        "arcadedb_oltp_language": arcadedb_oltp_language,
     }
 
 
@@ -7673,10 +7666,6 @@ def write_results(db_path: Path, args: argparse.Namespace, summary: dict):
         "query_result_hash_stable": summary.get("query_result_hash_stable"),
         "query_row_count_stable": summary.get("query_row_count_stable"),
         "benchmark_scope_note": summary.get("benchmark_scope_note"),
-        "arcadedb_oltp_language": summary.get(
-            "arcadedb_oltp_language",
-            args.arcadedb_oltp_language if args.db == "arcadedb" else None,
-        ),
     }
     results_path.parent.mkdir(parents=True, exist_ok=True)
     with open(results_path, "w", encoding="utf-8") as handle:
@@ -7757,7 +7746,7 @@ def run_in_docker(args) -> bool:
         filtered_args.append(arg)
 
     arcadedb_wheel_mount_path = None
-    if args.db == "arcadedb":
+    if args.db.startswith("arcadedb_"):
         wheel_candidates = sorted(
             (repo_root / "bindings/python/dist").glob("*embed*.whl")
         )
@@ -7845,15 +7834,16 @@ def main():
     parser.add_argument(
         "--db",
         choices=[
-            "arcadedb",
+            "arcadedb_sql",
+            "arcadedb_cypher",
             "ladybug",
             "ladybugdb",
             "graphqlite",
             "sqlite_native",
             "python_memory",
         ],
-        default="arcadedb",
-        help="Database to test (default: arcadedb)",
+        default="arcadedb_cypher",
+        help="Database to test (default: arcadedb_cypher)",
     )
     parser.add_argument(
         "--threads",
@@ -7908,12 +7898,6 @@ def main():
         action="store_true",
         help="For threads=1 only, assert final graph counts match deterministic single-thread baseline",
     )
-    parser.add_argument(
-        "--arcadedb-oltp-language",
-        choices=["cypher", "sql"],
-        default="cypher",
-        help="ArcadeDB OLTP CRUD language (default: cypher)",
-    )
     parser.add_argument(
         "--sqlite-profile",
         choices=SQLITE_PROFILE_CHOICES,
@@ -7928,8 +7912,6 @@ def main():
         parser.error("--jvm-heap-fraction must be > 0 and <= 1")
     if args.verify_single_thread_series and args.threads != 1:
         parser.error("--verify-single-thread-series requires --threads 1")
-    if args.db not in ("arcadedb",) and args.arcadedb_oltp_language != "cypher":
-        parser.error("--arcadedb-oltp-language is only applicable when --db arcadedb")
 
     ran = run_in_docker(args)
     if ran:
@@ -7940,7 +7922,7 @@ def main():
             args.mem_limit,
             args.jvm_heap_fraction,
         )
-        if args.db == "arcadedb"
+        if args.db.startswith("arcadedb_")
         else args.mem_limit
     )
     args.heap_size_effective = heap_size
@@ -7967,14 +7949,14 @@ def main():
     print(f"Threads: {args.threads}")
     print(f"Operations: {args.transactions:,}")
     print(f"Batch size: {args.batch_size}")
-    if args.db == "arcadedb":
+    if args.db.startswith("arcadedb_"):
         print(f"JVM heap size: {heap_size}")
-        print(f"ArcadeDB OLTP language: {args.arcadedb_oltp_language}")
+        print(f"ArcadeDB OLTP language: {args.db.removeprefix('arcadedb_')}")
     print(f"DB path: {db_path}")
     print(BENCHMARK_SCOPE_NOTE)
     print()
 
-    if args.db == "arcadedb":
+    if args.db.startswith("arcadedb_"):
         summary = run_graph_oltp_arcadedb(
             db_path=db_path,
             data_dir=data_dir,
@@ -7983,7 +7965,7 @@ def main():
             threads=args.threads,
             seed=args.seed,
             jvm_kwargs=jvm_kwargs,
-            oltp_language=args.arcadedb_oltp_language,
+            oltp_language=args.db.removeprefix("arcadedb_"),
         )
     elif args.db in ("ladybug", "ladybugdb"):
         summary = run_graph_oltp_ladybug(

bindings/python/examples/10_stackoverflow_graph_olap.py

@@ -4799,7 +4799,7 @@ def write_results(db_path: Path, args: argparse.Namespace, summary: dict):
             "arcadedb_olap_language",
             (
                 args.arcadedb_olap_language
-                if args.db in ("arcadedb", "arcadedb_cypher")
+                if args.db in ("arcadedb_sql", "arcadedb_cypher")
                 else None
             ),
         ),
@@ -4945,7 +4945,7 @@ def run_in_docker(args) -> bool:
         filtered_args.append(arg)
 
     arcadedb_wheel_mount_path = None
-    if args.db in ("arcadedb", "arcadedb_cypher"):
+    if args.db in ("arcadedb_sql", "arcadedb_cypher"):
         wheel_candidates = sorted(
             (repo_root / "bindings/python/dist").glob("*embed*.whl")
         )
@@ -5033,16 +5033,16 @@ def main():
     parser.add_argument(
         "--db",
         choices=[
-            "arcadedb",
+            "arcadedb_sql",
             "arcadedb_cypher",
             "ladybug",
             "ladybugdb",
             "graphqlite",
             "sqlite_native",
             "python_memory",
         ],
-        default="arcadedb",
-        help="Database to test (default: arcadedb)",
+        default="arcadedb_cypher",
+        help="Database to test (default: arcadedb_cypher)",
     )
     parser.add_argument(
         "--batch-size",
@@ -5122,7 +5122,9 @@ def main():
     if args.jvm_heap_fraction <= 0 or args.jvm_heap_fraction > 1:
         parser.error("--jvm-heap-fraction must be > 0 and <= 1")
 
-    args.arcadedb_olap_language = "cypher"
+    args.arcadedb_olap_language = None
+    if args.db.startswith("arcadedb_"):
+        args.arcadedb_olap_language = args.db.removeprefix("arcadedb_")
 
     ran = run_in_docker(args)
     if ran:
@@ -5133,7 +5135,7 @@ def main():
             args.mem_limit,
             args.jvm_heap_fraction,
         )
-        if args.db in ("arcadedb", "arcadedb_cypher")
+        if args.db in ("arcadedb_sql", "arcadedb_cypher")
         else args.mem_limit
     )
     args.heap_size_effective = heap_size
@@ -5155,7 +5157,7 @@ def main():
     print("=" * 80)
     print(f"Dataset: {args.dataset}")
     print(f"DB: {args.db}")
-    if args.db in ("arcadedb", "arcadedb_cypher"):
+    if args.db in ("arcadedb_sql", "arcadedb_cypher"):
         print(f"ArcadeDB OLAP language: {args.arcadedb_olap_language}")
     if args.db in ("sqlite_native", "graphqlite", "python_memory"):
         print(f"SQLite profile: {args.sqlite_profile}")
@@ -5171,7 +5173,7 @@ def main():
     stop_event, rss_state, rss_thread = start_rss_sampler()
     start_time = time.perf_counter()
 
-    if args.db in ("arcadedb", "arcadedb_cypher"):
+    if args.db in ("arcadedb_sql", "arcadedb_cypher"):
         summary = run_olap_arcadedb(
             db_path=db_path,
             data_dir=data_dir,

bindings/python/examples/11_vector_index_build.py

@@ -1421,7 +1421,7 @@ def run_in_docker(args) -> bool:
             filtered_args.extend(["--qdrant-host", "host.docker.internal"])
 
     arcadedb_wheel_mount_path = None
-    if args.backend == "arcadedb":
+    if args.backend == "arcadedb_sql":
         wheel_candidates = sorted(
             (repo_root / "bindings/python/dist").glob("*embed*.whl")
         )
@@ -1443,7 +1443,7 @@ def run_in_docker(args) -> bool:
     if args.backend == "pgvector" and image == "python:3.12-slim":
         image = default_docker_image(args.backend)
 
-    if args.backend == "arcadedb":
+    if args.backend == "arcadedb_sql":
         inner_cmd = " && ".join(
             [
                 "python -m venv /tmp/bench-venv",
@@ -1618,9 +1618,9 @@ def main() -> None:
     )
     parser.add_argument(
         "--backend",
-        choices=["arcadedb", "pgvector", "qdrant", "milvus", "faiss", "lancedb"],
-        default="arcadedb",
-        help="Vector backend (default: arcadedb)",
+        choices=["arcadedb_sql", "pgvector", "qdrant", "milvus", "faiss", "lancedb"],
+        default="arcadedb_sql",
+        help="Vector backend (default: arcadedb_sql)",
     )
     parser.add_argument(
         "--dataset",
@@ -1820,7 +1820,7 @@ def record(name: str, result, dur: float, rss_start: float, rss_end: float):
 
     runtime_versions: dict[str, str | None] = {}
 
-    if args.backend == "arcadedb":
+    if args.backend == "arcadedb_sql":
         stop_cpu = start_cpu_logger(2)
         import arcadedb_embedded as arcadedb
 
@@ -2307,7 +2307,7 @@ def server_pid_provider() -> int | None:
         },
         "environment": collect_runtime_metadata(
             args,
-            heap_size=(arcadedb_heap_size if args.backend == "arcadedb" else None),
+            heap_size=(arcadedb_heap_size if args.backend == "arcadedb_sql" else None),
             runtime_versions=runtime_versions,
         ),
         "budget": budget_allocation_report(args),
@@ -2346,7 +2346,7 @@ def server_pid_provider() -> int | None:
             },
             "arcadedb": {
                 "heap_size_effective": (
-                    arcadedb_heap_size if args.backend == "arcadedb" else None
+                    arcadedb_heap_size if args.backend == "arcadedb_sql" else None
                 ),
             },
             "qdrant": {