@tzwel/perftime

⏱️ perfTime

PerfTime is a simple and flexible utility for measuring function execution time. No depenencies.

It works by wrapping the performance.now() method and gives you a simple to use API for quick debugging purposes

Installation:

npm i @tzwel/perftime

Then require it in your project

const perfTime = require('@tzwel/perftime')

Usage:

One-liner

You can pass your function to perfTime and then call the .run() method. This exectues the function, measures its execution time and then logs the result.

new perfTime({function: yourFunctionHere}).run()
new perfTime({function: asyncFunction}).runAsync() // for asynchronous functions

Basic usage

function someRandomFunction() {
	// initialize perfTime and start measuring
	const measurement = new perfTime('function name (or you can leave this blank)').start()

	// code to be measured goes here

	measurement.stop() // stop the measurement and log results
	// => Executing 'someRandomFunction' took 0.006400000000000735ms
}

You can also start the measurement at any point later:

const measurement = new perfTime({function: someRandomFunction})
measurement.start()
// code to be measured goes here
measurement.stop()

The name of the measured function can be set in four different ways;

PerfTime accepts an options object or a string with a name as argument

new perfTime('some function name') // string name
new perfTime({functionName: 'some function name'}) // options object with provided function name
new perfTime({function: someFunction}) // options object with the function name derived automatically from the passed function
new perfTime() // empty, function gets called an *unnamed function*

Multiple measures and average

You can measure a function multiple times and then get its average execution time

[!WARNING] When getting average time, the first result always gets omitted because the first function call is always slower before optimizations take place

It can be done using the one-liner like this:

const measurement = new perfTime({function: someRandomFunction}).run(5) // benchmarks the function 5 times
console.log(measurement.averageTime) // log the average time of execution

// Async variant:
new perfTime({function: asyncFunction}).runAsync(5).then((measurement)=> {
	console.log(measurement.averageTime);
})

Or inside the function like this:

function someRandomFunction() {
	const measurement = new perfTime({function: someRandomFunction})
	
	for (let index = 0; index < 15; index++) { // a for loop that executes code multiple times
		measurement.start()
		// function code to be measured goes here
		measurement.stop()
	}

	console.log(measurement.averageTime);
}

Or by wrapping the function call in a for loop (you can implement this however you want)

const measurement = new perfTime({function: someRandomFunction})
for (let index = 0; index < 15; index++) {
	measurement.start()
	someRandomFunction()
	measurement.stop()
}
console.log(measurement.averageTime);

Async

Currently async functions can only be benchmarked using runAsync and are run one by one, not in parallel