From 401a63f54465ad82fbcdbdace8d08c13354125e5 Mon Sep 17 00:00:00 2001
From: Edward Langley <el-github@elangley.org>
Date: Sun, 5 Apr 2026 01:38:45 -0700
Subject: [PATCH] bench: add profiling workload generator

Generates Forth-style command scripts that build a multi-dimensional
model and exercise the grid aggregation hot path via repeated
export-csv calls. Used for profiling with samply.

Usage:
  python3 bench/gen_workload.py --scale 5 > /tmp/workload.txt
  cargo build --profile profiling
  samply record ./target/profiling/improvise script /tmp/workload.txt

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---
 bench/gen_workload.py | 83 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 83 insertions(+)
 create mode 100644 bench/gen_workload.py

diff --git a/bench/gen_workload.py b/bench/gen_workload.py
new file mode 100644
index 0000000..7391fc4
--- /dev/null
+++ b/bench/gen_workload.py
@@ -0,0 +1,83 @@
+#!/usr/bin/env python3
+"""Generate a profiling workload script for improvise.
+
+Usage:
+    python3 bench/gen_workload.py [--scale N] > bench/large_workload.txt
+    cargo build --release
+    time ./target/release/improvise script bench/large_workload.txt
+
+For flamegraph profiling:
+    samply record ./target/release/improvise script bench/large_workload.txt
+"""
+import argparse
+import random
+
+parser = argparse.ArgumentParser()
+parser.add_argument("--scale", type=int, default=1,
+                    help="Scale factor (1=small, 5=medium, 10=large)")
+parser.add_argument("--density", type=float, default=0.3,
+                    help="Cell density (0.0-1.0)")
+parser.add_argument("--exports", type=int, default=0,
+                    help="Number of export passes (0 = one per month)")
+args = parser.parse_args()
+
+random.seed(42)
+S = args.scale
+
+n_regions = 5 * S
+n_products = 8 * S
+n_months = 12
+n_channels = 4 + S
+measures = ["Revenue", "Cost", "Units"]
+
+regions = [f"R{i:03d}" for i in range(n_regions)]
+products = [f"P{i:03d}" for i in range(n_products)]
+months = [f"M{i:02d}" for i in range(1, n_months + 1)]
+channels = [f"Ch{i:02d}" for i in range(n_channels)]
+
+potential = n_regions * n_products * n_months * n_channels * len(measures)
+print(f"# Scale={S}, Density={args.density}")
+print(f"# {n_regions} regions × {n_products} products × {n_months} months × {n_channels} channels × {len(measures)} measures")
+print(f"# Potential cells: {potential}, Expected: ~{int(potential * args.density)}")
+print()
+
+for cat in ["Region", "Product", "Month", "Channel", "Measure"]:
+    print(f"add-category {cat}")
+
+for items, cat in [(regions, "Region"), (products, "Product"),
+                    (months, "Month"), (channels, "Channel"),
+                    (measures, "Measure")]:
+    for item in items:
+        print(f"add-item {cat} {item}")
+
+print("set-axis Region row")
+print("set-axis Product column")
+print("set-axis Month page")
+print("set-axis Channel none")
+print("set-axis Measure none")
+
+n = 0
+for r in regions:
+    for p in products:
+        for m in months:
+            for c in channels:
+                if random.random() < args.density:
+                    rev = random.randint(100, 10000)
+                    cost = random.randint(50, rev)
+                    units = random.randint(1, 500)
+                    print(f"set-cell {rev} Region/{r} Product/{p} Month/{m} Channel/{c} Measure/Revenue")
+                    print(f"set-cell {cost} Region/{r} Product/{p} Month/{m} Channel/{c} Measure/Cost")
+                    print(f"set-cell {units} Region/{r} Product/{p} Month/{m} Channel/{c} Measure/Units")
+                    n += 3
+
+print(f"# Total cells: {n}")
+
+print('add-formula Measure "Profit = Revenue - Cost"')
+print('add-formula Measure "Margin = Profit / Revenue"')
+print('add-formula Measure "AvgPrice = Revenue / Units"')
+
+n_exports = args.exports if args.exports > 0 else n_months
+for i, m in enumerate(months[:n_exports]):
+    print(f"set-page Month {m} . export-csv /tmp/improvise_bench_{i:02d}.csv")
+
+print("# Done")