bench(py): track per-function data-processing metrics

README.md

@@ -54,6 +54,9 @@ cargo run -p openquant --example research_notebook_smoke
 
 # Python pipeline micro-benchmark (for speed demos)
 uv run --python .venv/bin/python python python/benchmarks/benchmark_pipeline.py --iterations 30 --bars 2048
+
+# Python data-processing benchmark (per-function throughput/latency + JSON output)
+uv run --python .venv/bin/python python python/benchmarks/benchmark_data_processing.py --rows-per-symbol 200000 --symbols 4 --iterations 7 --out benchmarks/data_processing/latest.json
 ```
 
 ## Crate Layout

benchmarks/data_processing/README.md

@@ -0,0 +1,38 @@
+# Data Processing Benchmarks
+
+This folder tracks Python-facing `openquant.data` performance metrics.
+
+## Files
+
+- `latest.json`: most recent benchmark run output.
+- `baseline.json`: pinned comparison baseline for regression checks.
+
+## Generate Latest Metrics
+
+```bash
+just py-bench-data
+```
+
+or directly:
+
+```bash
+uv run --python .venv/bin/python python python/benchmarks/benchmark_data_processing.py \
+  --rows-per-symbol 200000 \
+  --symbols 4 \
+  --iterations 7 \
+  --out benchmarks/data_processing/latest.json
+```
+
+## Compare Against Baseline
+
+```bash
+just py-bench-data-compare
+```
+
+## Refresh Baseline
+
+After reviewing and accepting new performance:
+
+```bash
+cp benchmarks/data_processing/latest.json benchmarks/data_processing/baseline.json
+```

benchmarks/data_processing/latest.json

@@ -0,0 +1,75 @@
+{
+  "dataset": {
+    "rows_per_symbol": 100000,
+    "symbols": 4,
+    "total_rows": 400000,
+    "estimated_bytes": 20800000
+  },
+  "iterations": 5,
+  "functions": [
+    {
+      "name": "clean_ohlcv",
+      "rows": 400000,
+      "bytes_estimate": 20800000,
+      "iterations": 5,
+      "mean_ms": 431.8176897984813,
+      "p50_ms": 427.49624899443006,
+      "p95_ms": 430.2287489990704,
+      "std_ms": 8.461222812078079,
+      "rows_per_sec": 926316.8449321059,
+      "mb_per_sec": 48.168475936469505
+    },
+    {
+      "name": "data_quality_report",
+      "rows": 400000,
+      "bytes_estimate": 20800000,
+      "iterations": 5,
+      "mean_ms": 202.2392454004148,
+      "p50_ms": 202.07314000435872,
+      "p95_ms": 205.02412800124148,
+      "std_ms": 2.602492864446543,
+      "rows_per_sec": 1977855.481056792,
+      "mb_per_sec": 102.84848501495318
+    },
+    {
+      "name": "align_calendar",
+      "rows": 400000,
+      "bytes_estimate": 24000000,
+      "iterations": 5,
+      "mean_ms": 487.6979159977054,
+      "p50_ms": 484.19221499352716,
+      "p95_ms": 485.0531679985579,
+      "std_ms": 8.35346004858996,
+      "rows_per_sec": 820179.8426423499,
+      "mb_per_sec": 49.21079055854099
+    },
+    {
+      "name": "load_ohlcv_csv",
+      "rows": 400000,
+      "bytes_estimate": 28228821,
+      "iterations": 5,
+      "mean_ms": 234.64987420011312,
+      "p50_ms": 232.7873370013549,
+      "p95_ms": 235.60187700059032,
+      "std_ms": 6.00703020230825,
+      "rows_per_sec": 1704667.4385125546,
+      "mb_per_sec": 120.30187996574853
+    },
+    {
+      "name": "load_ohlcv_parquet",
+      "rows": 400000,
+      "bytes_estimate": 2243334,
+      "iterations": 5,
+      "mean_ms": 458.6696180005674,
+      "p50_ms": 461.2510000006296,
+      "p95_ms": 468.5403839976061,
+      "std_ms": 11.962792657264071,
+      "rows_per_sec": 872087.4117271599,
+      "mb_per_sec": 4.890958354248841
+    }
+  ],
+  "tables": {
+    "core": "name                          rows     mean_ms      p95_ms        rows/s        MB/s\n------------------------------------------------------------------------------------\nclean_ohlcv                400,000      431.82      430.23     926.32K/s        48.2\ndata_quality_report        400,000      202.24      205.02       1.98M/s       102.8\nalign_calendar             400,000      487.70      485.05     820.18K/s        49.2",
+    "io": "name                          rows     mean_ms      p95_ms        rows/s        MB/s\n------------------------------------------------------------------------------------\nload_ohlcv_csv             400,000      234.65      235.60       1.70M/s       120.3\nload_ohlcv_parquet         400,000      458.67      468.54     872.09K/s         4.9"
+  }
+}

docs/python_bindings.md

@@ -22,6 +22,7 @@ Quick performance showcase:
 
 ```bash
 uv run --python .venv/bin/python python python/benchmarks/benchmark_pipeline.py --iterations 30 --bars 2048
+uv run --python .venv/bin/python python python/benchmarks/benchmark_data_processing.py --rows-per-symbol 200000 --symbols 4 --iterations 7 --out benchmarks/data_processing/latest.json
 ```
 
 Build a wheel:

justfile

@@ -65,6 +65,12 @@ py-setup:
 py-bench:
     uv run --python .venv/bin/python python python/benchmarks/benchmark_pipeline.py --iterations 30 --bars 2048
 
+py-bench-data:
+    uv run --python .venv/bin/python python python/benchmarks/benchmark_data_processing.py --rows-per-symbol 200000 --symbols 4 --iterations 7 --out benchmarks/data_processing/latest.json
+
+py-bench-data-compare:
+    uv run --python .venv/bin/python python python/benchmarks/benchmark_data_processing.py --rows-per-symbol 200000 --symbols 4 --iterations 7 --out benchmarks/data_processing/latest.json --baseline benchmarks/data_processing/baseline.json
+
 exp-run:
     uv run --python .venv/bin/python python experiments/run_pipeline.py --config experiments/configs/futures_oil_baseline.toml --out experiments/artifacts