bench: new files

bench/interp-dispatch.js

@@ -0,0 +1,18 @@
+// taken mostly from hermes
+
+function bench (lc, fc) {
+  var n, fact;
+  var res = 0;
+  while (--lc >= 0) {
+      n = fc;
+      fact = n;
+      while (--n > 1)
+          fact *= n;
+      res += fact;
+  }
+  return res;
+}
+
+let t1 = performance.now();
+var res = bench(4e6, 100);
+print(performance.now() - t1);

bench/pearson.js

@@ -0,0 +1,41 @@
+const start = performance.now();
+
+// Define your two arrays of numbers
+const array1 = [1, 2, 3, 4, 5];
+const array2 = [5, 4, 3, 2, 1];
+
+const mean = arr => {
+  arr ??= [];
+
+  let sum = 0;
+  for (const x of arr) {
+    sum += x;
+  }
+
+  return sum / arr.length;
+};
+
+// Calculate the mean of each array
+// const mean1 = array1.reduce((acc, val) => acc + val, 0) / array1.length;
+// const mean2 = array2.reduce((acc, val) => acc + val, 0) / array2.length;
+const mean1 = mean(array1);
+const mean2 = mean(array2);
+
+// Calculate the numerator and denominators for Pearson correlation
+let numerator = 0;
+let denominator1 = 0;
+let denominator2 = 0;
+
+for (let i = 0; i < array1.length; i++) {
+  const diff1 = array1[i] - mean1;
+  const diff2 = array2[i] - mean2;
+  numerator += diff1 * diff2;
+  denominator1 += diff1 * diff1;
+  denominator2 += diff2 * diff2;
+}
+
+// Calculate the correlation coefficient
+const correlationCoefficient = numerator / Math.sqrt(denominator1 * denominator2);
+
+console.log(correlationCoefficient);
+console.log(performance.now() - start);

bench/wellformed.js

@@ -0,0 +1,12 @@
+let t = performance.now();
+
+let i = 0;
+while (i < 100_000) {
+  'abc'.isWellFormed() // true
+  '🔥'.isWellFormed() // true
+  '\uD83D'.isWellFormed() // false
+
+  i += 1;
+}
+
+console.log(performance.now() - t);