[Autoloop: perf-comparison] Iteration 409: readStata/toStata benchmark

Add TypeScript (readStata from in-memory DTA buffer) and Python (pd.read_stata from temp file) benchmarks for 10k-row DataFrame with int32/float64/string columns, 20 measured iterations.

Run: https://github.com/githubnext/tsb/actions/runs/29668492326

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: github-actions[bot]

benchmarks/pandas/bench_readStata.py

@@ -0,0 +1,37 @@
+"""Benchmark: readStata / toStata round-trip on a 10k-row DataFrame"""
+import json, time, tempfile, os
+import numpy as np
+import pandas as pd
+
+ROWS = 10_000
+WARMUP = 3
+ITERATIONS = 20
+
+ids = np.arange(ROWS, dtype=np.int32)
+values = np.sin(np.arange(ROWS) * 0.01) * 1000
+categories = np.array([f"cat_{i % 5}" for i in range(ROWS)])
+
+df = pd.DataFrame({"id": ids, "value": values, "category": categories})
+
+# Write to a temp Stata file so read_stata benchmarks read from disk
+tmp = tempfile.NamedTemporaryFile(suffix=".dta", delete=False)
+tmp.close()
+df.to_stata(tmp.name, write_index=False)
+
+# Warm up
+for _ in range(WARMUP):
+    pd.read_stata(tmp.name)
+
+start = time.perf_counter()
+for _ in range(ITERATIONS):
+    pd.read_stata(tmp.name)
+total = (time.perf_counter() - start) * 1000
+
+os.unlink(tmp.name)
+
+print(json.dumps({
+    "function": "readStata",
+    "mean_ms": total / ITERATIONS,
+    "iterations": ITERATIONS,
+    "total_ms": total,
+}))

benchmarks/tsb/bench_readStata.ts

@@ -0,0 +1,42 @@
+/**
+ * Benchmark: readStata / toStata round-trip on a 10k-row DataFrame
+ */
+import { DataFrame, Series, readStata, toStata } from "../../src/index.js";
+
+const ROWS = 10_000;
+const WARMUP = 3;
+const ITERATIONS = 20;
+
+// Build a DataFrame with numeric and string columns
+const ids = Int32Array.from({ length: ROWS }, (_, i) => i);
+const values = Float64Array.from({ length: ROWS }, (_, i) => Math.sin(i * 0.01) * 1000);
+const categories = Array.from({ length: ROWS }, (_, i) => `cat_${i % 5}`);
+
+const df = new DataFrame({
+  id: new Series(ids),
+  value: new Series(values),
+  category: new Series(categories),
+});
+
+// Serialize once so readStata benchmarks read from a pre-built buffer
+const buf = toStata(df);
+
+// Warm up
+for (let i = 0; i < WARMUP; i++) {
+  readStata(buf);
+}
+
+const start = performance.now();
+for (let i = 0; i < ITERATIONS; i++) {
+  readStata(buf);
+}
+const total = performance.now() - start;
+
+console.log(
+  JSON.stringify({
+    function: "readStata",
+    mean_ms: total / ITERATIONS,
+    iterations: ITERATIONS,
+    total_ms: total,
+  }),
+);