@curl-runner/cli 1.13.0 → 1.15.0

package/src/utils/stats.ts

@@ -0,0 +1,151 @@
+import type { ProfileStats } from '../types/config';
+
+/**
+ * Calculate percentile from sorted array.
+ * Uses linear interpolation for non-integer indices.
+ */
+export function calculatePercentile(sorted: number[], percentile: number): number {
+  if (sorted.length === 0) {
+    return 0;
+  }
+  if (sorted.length === 1) {
+    return sorted[0];
+  }
+
+  const index = (percentile / 100) * (sorted.length - 1);
+  const lower = Math.floor(index);
+  const upper = Math.ceil(index);
+  const fraction = index - lower;
+
+  if (lower === upper) {
+    return sorted[lower];
+  }
+  return sorted[lower] * (1 - fraction) + sorted[upper] * fraction;
+}
+
+/**
+ * Calculate arithmetic mean.
+ */
+export function calculateMean(values: number[]): number {
+  if (values.length === 0) {
+    return 0;
+  }
+  return values.reduce((sum, v) => sum + v, 0) / values.length;
+}
+
+/**
+ * Calculate standard deviation.
+ */
+export function calculateStdDev(values: number[], mean: number): number {
+  if (values.length <= 1) {
+    return 0;
+  }
+  const squaredDiffs = values.map((v) => (v - mean) ** 2);
+  const variance = squaredDiffs.reduce((sum, v) => sum + v, 0) / values.length;
+  return Math.sqrt(variance);
+}
+
+/**
+ * Calculate profile statistics from raw timings.
+ */
+export function calculateProfileStats(
+  timings: number[],
+  warmup: number,
+  failures: number,
+): ProfileStats {
+  // Exclude warmup iterations
+  const effectiveTimings = timings.slice(warmup);
+  const sorted = [...effectiveTimings].sort((a, b) => a - b);
+
+  const mean = calculateMean(sorted);
+  const totalIterations = timings.length;
+  const effectiveIterations = effectiveTimings.length;
+
+  return {
+    iterations: effectiveIterations,
+    warmup,
+    min: sorted.length > 0 ? sorted[0] : 0,
+    max: sorted.length > 0 ? sorted[sorted.length - 1] : 0,
+    mean: Math.round(mean * 100) / 100,
+    median: Math.round(calculatePercentile(sorted, 50) * 100) / 100,
+    p50: Math.round(calculatePercentile(sorted, 50) * 100) / 100,
+    p95: Math.round(calculatePercentile(sorted, 95) * 100) / 100,
+    p99: Math.round(calculatePercentile(sorted, 99) * 100) / 100,
+    stdDev: Math.round(calculateStdDev(sorted, mean) * 100) / 100,
+    failures,
+    failureRate: totalIterations > 0 ? Math.round((failures / totalIterations) * 10000) / 100 : 0,
+    timings: effectiveTimings,
+  };
+}
+
+/**
+ * Generate ASCII histogram for latency distribution.
+ */
+export function generateHistogram(timings: number[], buckets = 10, width = 40): string[] {
+  if (timings.length === 0) {
+    return ['No data'];
+  }
+
+  const min = Math.min(...timings);
+  const max = Math.max(...timings);
+  const range = max - min || 1;
+  const bucketSize = range / buckets;
+
+  // Count values per bucket
+  const counts = new Array(buckets).fill(0);
+  for (const t of timings) {
+    const bucket = Math.min(Math.floor((t - min) / bucketSize), buckets - 1);
+    counts[bucket]++;
+  }
+
+  const maxCount = Math.max(...counts);
+  const lines: string[] = [];
+
+  for (let i = 0; i < buckets; i++) {
+    const bucketMin = min + i * bucketSize;
+    const bucketMax = min + (i + 1) * bucketSize;
+    const barLength = maxCount > 0 ? Math.round((counts[i] / maxCount) * width) : 0;
+    const bar = '█'.repeat(barLength);
+    const label = `${bucketMin.toFixed(0).padStart(6)}ms - ${bucketMax.toFixed(0).padStart(6)}ms`;
+    lines.push(`${label} │${bar} ${counts[i]}`);
+  }
+
+  return lines;
+}
+
+/**
+ * Export stats to CSV format.
+ */
+export function exportToCSV(stats: ProfileStats, _requestName: string): string {
+  const headers = ['iteration', 'latency_ms'];
+  const rows = stats.timings.map((t, i) => `${i + 1},${t}`);
+  return [headers.join(','), ...rows].join('\n');
+}
+
+/**
+ * Export stats to JSON format.
+ */
+export function exportToJSON(stats: ProfileStats, requestName: string): string {
+  return JSON.stringify(
+    {
+      request: requestName,
+      summary: {
+        iterations: stats.iterations,
+        warmup: stats.warmup,
+        failures: stats.failures,
+        failureRate: stats.failureRate,
+        min: stats.min,
+        max: stats.max,
+        mean: stats.mean,
+        median: stats.median,
+        p50: stats.p50,
+        p95: stats.p95,
+        p99: stats.p99,
+        stdDev: stats.stdDev,
+      },
+      timings: stats.timings,
+    },
+    null,
+    2,
+  );
+}