Add timing benchmark wrapper script

Add run_timing_benchmark.sh script to automate timing benchmarks
for simulation campaigns with configurable parameters.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <[email protected]>

Author: Sakib Rahman

scripts/run_timing_benchmark.sh

@@ -0,0 +1,101 @@
+#!/bin/bash
+
+# Wrapper script to run timing benchmark for a specific dataset locally
+# Usage: ./run_timing_benchmark.sh <path-to-dataset.csv>
+
+set -e
+
+# Determine script directory and repository root
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+REPO_ROOT="$(cd "$SCRIPT_DIR/.." && pwd)"
+
+if [ $# -eq 0 ]; then
+    echo "ERROR: No dataset CSV file provided"
+    echo "Usage: $0 <path-to-dataset.csv>"
+    echo ""
+    echo "Examples:"
+    echo "  $0 SIDIS/DIJET/pythia6.428-dijet-v1.0_PGF_noRC_ep_18x275_q2_1to20000_ab.csv"
+    echo "  $0 EXCLUSIVE/UPSILON.csv"
+    exit 1
+fi
+
+DATA="$1"
+
+# Change to repo root directory
+cd "$REPO_ROOT"
+
+if [ ! -f "$DATA" ]; then
+    echo "ERROR: File '$DATA' not found"
+    exit 1
+fi
+
+# Set default environment variables (can be overridden)
+export DETECTOR_CONFIG="${DETECTOR_CONFIG:-epic_craterlake}"
+export DETECTOR_VERSION="${DETECTOR_VERSION:-main}"
+export EBEAM="${EBEAM:-18}"
+export PBEAM="${PBEAM:-275}"
+export NEVENTS_PER_TEST="${NEVENTS_PER_TEST:-100}"
+export IMAGE_TAG="${IMAGE_TAG:-nightly}"
+export RESULTS_BASE="results/${IMAGE_TAG}/${DETECTOR_CONFIG}/${DETECTOR_VERSION}"
+
+# Set up parallel alias
+shopt -s expand_aliases
+alias parallel='parallel -k --lb -j 1 --colsep ","'
+
+echo "========================================"
+echo "Running timing benchmark for: $DATA"
+echo "========================================"
+echo "DETECTOR_CONFIG = ${DETECTOR_CONFIG}"
+echo "DETECTOR_VERSION = ${DETECTOR_VERSION}"
+echo "EBEAM = ${EBEAM}"
+echo "PBEAM = ${PBEAM}"
+echo "NEVENTS_PER_TEST = ${NEVENTS_PER_TEST}"
+echo "IMAGE_TAG = ${IMAGE_TAG}"
+echo "RESULTS_BASE = ${RESULTS_BASE}"
+echo "========================================"
+echo ""
+
+# Stage 1: Glob - Find matching files
+echo "[1/3] Running glob stage to find matching files..."
+mkdir -p $(dirname ${RESULTS_BASE}/datasets/glob/$DATA)
+grep -v "^\#" $DATA | parallel scripts/glob.sh ${RESULTS_BASE}/datasets/glob/$DATA {}
+sort -o ${RESULTS_BASE}/datasets/glob/$DATA ${RESULTS_BASE}/datasets/glob/$DATA
+echo "✓ Glob stage complete"
+echo ""
+
+# Stage 2: Nevents - Count events
+echo "[2/3] Running nevents stage to count events..."
+mkdir -p $(dirname ${RESULTS_BASE}/datasets/nevents/$DATA)
+grep -v "^\#" ${RESULTS_BASE}/datasets/glob/$DATA | parallel scripts/count_events.sh ${RESULTS_BASE}/datasets/nevents/$DATA {}
+sort -o ${RESULTS_BASE}/datasets/nevents/$DATA ${RESULTS_BASE}/datasets/nevents/$DATA
+echo "✓ Nevents stage complete"
+echo ""
+
+# Stage 3: Timings - Run timing benchmarks
+echo "[3/3] Running timings stage (this may take a while)..."
+mkdir -p $(dirname ${RESULTS_BASE}/datasets/timings/$DATA)
+# First file with timing measurement
+grep -v "^\#" ${RESULTS_BASE}/datasets/nevents/$DATA | sed '1!d' | parallel scripts/determine_timing.sh ${RESULTS_BASE}/datasets/timings/$DATA {}
+# Remaining files using timing from first file
+IFS="," read file ext nevents dt0 dt1 < ${RESULTS_BASE}/datasets/timings/$DATA
+export dt0 dt1
+grep -v "^\#" ${RESULTS_BASE}/datasets/nevents/$DATA | sed '1d' | parallel scripts/determine_timing.sh ${RESULTS_BASE}/datasets/timings/$DATA {}
+sort -o ${RESULTS_BASE}/datasets/timings/$DATA ${RESULTS_BASE}/datasets/timings/$DATA
+echo "✓ Timings stage complete"
+echo ""
+
+# Show results
+echo "========================================"
+echo "RESULTS"
+echo "========================================"
+echo ""
+echo "Timing results saved to:"
+echo "  ${RESULTS_BASE}/datasets/timings/$DATA"
+echo ""
+echo "Summary:"
+cat ${RESULTS_BASE}/datasets/timings/$DATA | awk 'BEGIN {FS=","} {sum+=$3*$5+$4} END {printf("  Total core-hours: %.2f\n",sum/3600)}'
+cat ${RESULTS_BASE}/datasets/timings/$DATA | awk 'BEGIN {FS=","} {sum+=$3*$7+$6} END {printf("  Total size (full): %.2f GB\n",sum/1048576)}'
+cat ${RESULTS_BASE}/datasets/timings/$DATA | awk 'BEGIN {FS=","} {sum+=$3*$9+$8} END {printf("  Total size (reco): %.2f GB\n",sum/1048576)}'
+echo ""
+echo "Full timing data:"
+cat ${RESULTS_BASE}/datasets/timings/$DATA