Add black-box cut optimisation tool to extras

extras/benchmark/pyproject.toml

@@ -3,7 +3,7 @@
 # SPDX-License-Identifier: MPL-2.0
 
 [project]
-name = ""
+name = "traccc-bench-tools"
 version = "0.0.1"
 requires-python = "~=3.9"
 dependencies = [
@@ -15,9 +15,6 @@ dependencies = [
 [dependency-groups]
 dev = ["black>=25.1.0,<26"]
 
-[tool.uv]
-package = false
-
 [build-system]
 requires = ["hatchling"]
 build-backend = "hatchling.build"

extras/cut_optimiser/.gitignore

@@ -0,0 +1,5 @@
+*.csv
+*.csv.bak
+*.json
+.venv/
+!example.json

extras/cut_optimiser/example.json

@@ -0,0 +1,18 @@
+{
+    "input": {
+        "event_dir": "/data/Acts/odd-simulations-20240506/geant4_ttbar_mu200",
+        "digitization_file": "geometries/odd/odd-digi-geometric-config.json",
+        "detector_file": "geometries/odd/odd-detray_geometry_detray.json",
+        "grid_file": "geometries/odd/odd-detray_surface_grids_detray.json",
+        "material_file": "geometries/odd/odd-detray_material_detray.json"
+    },
+    "config": {
+        "truth-finding-min-track-candidates": 7,
+        "track-candidates-range": "7:100"
+    },
+    "parameters": {
+        "max-num-branches-per-surface": [1, 2, 3],
+        "max-num-skipping-per-cand": [2, 3],
+        "chi2-max": [2.5, 5.0, 7.5, 10.0, 15.0, 20.0]
+    }
+}

extras/cut_optimiser/find_pareto_set.py

@@ -0,0 +1,97 @@
+import argparse
+import csv
+import logging
+import pathlib
+
+
+log = logging.getLogger("find_pareto_set")
+
+
+def main():
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument(
+        "db",
+        type=pathlib.Path,
+        help="the CSV database file",
+    )
+
+    parser.add_argument(
+        "-v",
+        "--verbose",
+        help="enable verbose output",
+        action="store_true",
+    )
+
+    args = parser.parse_args()
+
+    logging.basicConfig(
+        level=logging.DEBUG if (args.verbose or False) else logging.INFO,
+        format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
+    )
+
+    results = []
+
+    total_results = 0
+
+    with open(args.db, "r") as f:
+        reader = csv.DictReader(f)
+        for i in reader:
+            total_results += 1
+            if i["success"] != "0":
+                results.append({k: float(v) for k, v in i.items()})
+
+    log.info(
+        "Database contained %d results of which %d are valid",
+        total_results,
+        len(results),
+    )
+
+    pareto_set = []
+
+    for i, m in enumerate(results):
+        for j, n in enumerate(results):
+            if i == j:
+                continue
+
+            if (
+                n["rec_throughput"] <= m["rec_throughput"]
+                and n["efficiency"] >= m["efficiency"]
+                and n["fake_rate"] <= m["fake_rate"]
+                and n["duplicate_rate"] <= m["duplicate_rate"]
+            ):
+                log.debug(
+                    "Removing %s from the Pareto set because %s is superior",
+                    str(n),
+                    str(m),
+                )
+                break
+        else:
+            pareto_set.append(m)
+
+    log.info("Pareto set contains %d elements:", len(pareto_set))
+
+    for i in sorted(pareto_set, key=lambda x: x["rec_throughput"], reverse=True):
+        log.info(
+            "  Eff. %.2f, fake rate %.2f, duplicate rate %.2f with reciprocal througput %.1fms is achieved by setup {%s}",
+            100.0 * i["efficiency"],
+            i["fake_rate"],
+            i["duplicate_rate"],
+            i["rec_throughput"] * 1000.0,
+            ", ".join(
+                "%s: %s" % (k, str(v))
+                for k, v in i.items()
+                if k
+                not in [
+                    "efficiency",
+                    "fake_rate",
+                    "duplicate_rate",
+                    "rec_throughput",
+                    "success",
+                ]
+            ),
+        )
+
+
+if __name__ == "__main__":
+    main()