tatami-ng

tatami-ng

Cross JavaScript runtime benchmarking library

Install

Node

npm install tatami-ng

Deno

deno add npm:tatami-ng

Deno versions >= 1.40.x are supported.

The --allow-hrtime permission flag is recommended to allow high-resolution time measurement.

Bun

bun add tatami-ng

Bun versions >= 1.x are supported.

Browser

<script type="module">
import {
 ...
} from 'https://cdn.jsdelivr.net/npm/tatami-ng@0.2.0/src/cli.mjs'
</script>

Example

import { baseline, bench, clear, group, run } from 'tatami-ng';

bench('noop', () => {});
bench('noop2', () => {});

group('group', () => {
  baseline('baseline', () => {});
  bench('Date.now()', () => Date.now());
  bench('performance.now()', () => performance.now());
});

group({ name: 'group2', summary: false }, () => {
  bench('new Array(0)', () => new Array(0));
  bench('new Array(1024)', () => new Array(1024));
});

await run({
  units: false, // print units cheatsheet (default: false)
  silent: false, // enable/disable stdout output (default: undefined)
  json: false, // enable/disable json output (default: false)
  colors: true, // enable/disable colors (default: true)
  samples: 128, // minimum number of benchmark samples (default: 128)
  iter: true, // enable/disable iter/s column (default: true)
  avg: true, // enable/disable time (avg) column (default: true)
  min_max: true, // enable/disable (min...max) column (default: true)
  percentiles: false, // enable/disable percentile columns (default: true)
});

clear();

Development

The JavaScript runtime environment used for development is Bun.

License

MIT © Evan, Jerome Benoit

Changelog

[0.2.0] - 2024-04-07

Fixed

• Fix measure() asynchronous implementation.
• Fix percentiles computation.
• Fix inconsistent benchmark behavior: the samples statistical significance shall be done on the same measurement base.

Changed

• Add clear() benchmark clearing primitive.
• Optimize JavaScript runtime environment detection code (constify environment variables).
• Add IQR samples outlier detection and removal.
• Add standard deviation to JSON report.
• Add samples: number to run() options specifying the minimum number of benchmark samples.
• Add options: { before: () => void | Promise<void>; after: () => void | Promise<void> } to bench() and baseline() specifying callbacks to run before and after the benchmark.
• Add options: { before: () => void | Promise<void>; after: () => void | Promise<void> } to group() specifying callbacks to run before and after the group.
• Add iter/s to report.
• Add package publication GitHub workflow.