Benchmark: lodash merge vs deepmerge 4.2.2 vs own merge implementation

HTML Preparation code:

<script src='https://cdnjs.cloudflare.com/ajax/libs/lodash.js/4.17.5/lodash.min.js'></script>

<script src='https://unpkg.com/deepmerge@4.2.2/dist/umd.js'></script>

Script Preparation code:

var a = { a: 'oh', b: 'my', c: { a: 'a', b: { c: 'c' } } };
var b = { c: { b: { d: 'a'  }, c: { d: 'd' } } };
var c = _.merge({}, a, b);

Tests:

lodash
var a = { a: 'oh', b: 'my', c: { a: 'a', b: { c: 'c' } } }; var b = { c: { b: { d: 'a' }, c: { d: 'd' } } }; var c = _.merge({}, a, b);
deepmerge
var a = { a: 'oh', b: 'my', c: { a: 'a', b: { c: 'c' } } }; var b = { c: { b: { d: 'a' }, c: { d: 'd' } } }; var c = deepmerge({}, a, b);
own impl
function isMergable(property) { return Array.isArray(property) || typeof property === 'object' } function propertyTypesDoMatch(propA, propB) { if (propA && propB) { return propA.constructor.name === propB.constructor.name } return false } const basics = ['string', 'number', 'boolean', 'bigint', 'undefined', 'symbol', 'function'] function deepEqual(i1, i2) { if (basics.includes(typeof i1) && typeof i1 === typeof i2) { return i1 === i2 } if (Array.isArray(i1) && Array.isArray(i2)) { return i1.every((subItem) => arrayContainsSame(i2, subItem)) } if (typeof i1 === typeof i2 && typeof i1 === 'object') { return Object.entries(i1).every(([key, value]) => { return deepEqual(value, i2[key]) }) } return false } function arrayContainsSame(items, item) { // skippable for loop makes it 1ms faster for (let c = 0; c < items.length; c++) { if (deepEqual(items[c], item)) { return true } } } function mergeArrays(array1, array2) { const all = [...array1] array2.forEach((i) => { if (!arrayContainsSame(all, i)) { all.push(i) } }) return all } function mergeConfigurations(rootConfig, additionalConfig) { let copiedConfig = { ...rootConfig } Object.keys(additionalConfig).forEach((key) => { const doTypesMatch = propertyTypesDoMatch(copiedConfig[key], additionalConfig[key]) const propExistsInRoot = Object.hasOwnProperty.call(copiedConfig, key) if (propExistsInRoot && isMergable(copiedConfig[key]) && doTypesMatch) { if (Array.isArray(copiedConfig[key])) { copiedConfig[key] = mergeArrays(copiedConfig[key], additionalConfig[key]) } else { copiedConfig[key] = mergeConfigurations(copiedConfig[key], additionalConfig[key]) } } else if (propExistsInRoot && doTypesMatch) { copiedConfig[key] = additionalConfig[key] } else { copiedConfig = Object.assign(copiedConfig, { [key]: additionalConfig[key] }) } }) return copiedConfig } var a = { a: 'oh', b: 'my', c: { a: 'a', b: { c: 'c' } } }; var b = { c: { b: { d: 'a' }, c: { d: 'd' } } }; var c = mergeConfigurations(a, b)

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results

Test case name	Result
lodash
deepmerge
own impl

Fastest: N/A

Slowest: N/A

Latest run results:

Run details: (Test run date: one year ago)

User agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36

Browser/OS: Chrome 135 on Windows

View result in a separate tab

Test name	Executions per second
lodash	309929.2 Ops/sec
deepmerge	708380.6 Ops/sec
own impl	1251469.4 Ops/sec

Autogenerated LLM Summary (model llama3.2:3b, generated one year ago):

LLMs can make mistakes. Check important info.

I'll explain the benchmarking setup and test cases in detail.

**Benchmark Definition:**

The benchmark measures the performance of three JavaScript libraries for merging objects:

1. `lodash`: Lodash is a popular utility library that provides a wide range of functions for working with data, including object manipulation. The `_merge` function is used to merge two or more objects into one.
2. `deepmerge`: Deepmerge is a lightweight library specifically designed for merging objects. It provides a simple and efficient way to merge objects without modifying the original objects.
3. The custom implementation (`own impl`): This is the internal implementation of the merge logic, written by the developer.

**Test Cases:**

The benchmark consists of three test cases:

1. `lodash`: Tests the performance of Lodash's `_merge` function with two example input objects.
2. `deepmerge`: Tests the performance of Deepmerge's merging function with two example input objects. The version used is 4.2.2.
3. `own impl`: Tests the performance of the custom implementation written by the developer, using the same example input objects as before.

**Options Compared:**

The benchmark compares the performance of each library in terms of the number of executions per second (ExecutionsPerSecond).

**Pros and Cons of Each Approach:**

1. **Lodash (`_merge` function)**:
	* Pros:
		+ Wide range of features and functionality.
		+ Well-established and widely used.
	* Cons:
		+ Larger size due to its extensive feature set.
		+ May be overkill for simple object merging tasks.
2. **Deepmerge**:
	* Pros:
		+ Lightweight and efficient.
		+ Specifically designed for merging objects.
	* Cons:
		+ Limited features compared to Lodash.
		+ May require additional setup for more complex use cases.
3. **Custom Implementation (`own impl`)**:
	* Pros:
		+ Tailored to the specific requirements of the benchmark.
		+ Can be optimized for performance and efficiency.
	* Cons:
		+ Requires significant development effort.
		+ May not be reusable in other contexts.

**Latest Benchmark Result:**

The latest results show that:

1. The custom implementation (`own impl`) performs best, with an average of 744316 executions per second across all browsers and devices.
2. Deepmerge follows closely, with an average of 493385 executions per second.
3. Lodash performs last, with an average of 336120 executions per second.

Overall, the benchmark suggests that the custom implementation provides the best performance for object merging tasks, followed closely by Deepmerge.

Related benchmarks:

lodash merge vs deepmerge 4.2.2 vs own merge implementation (version: 0)

Comparing performance of: lodash vs deepmerge vs own impl

Created: 3 years ago by: Guest

Jump to the latest result

lodash

deepmerge

own impl

Suite status: <idle, ready to run>

Fastest: N/A

Slowest: N/A

Autogenerated LLM Summary (model llama3.2:3b, generated one year ago):