Benchmark: Map vs filter for unique properties in object

Script Preparation code:

const uniqueItems = {};

const arr = [{
        place: "here",
        name: "x",
        other: "other stuff1"
    },
    {
        place: "there",
        name: "x",
        other: "other stuff2"
    },
    {
        place: "here",
        name: "y",
        other: "other stuff4"
    },
    {
        place: "here",
        name: "z",
        other: "other stuff5"
    }
]

function getUniqueListByFilter() {
    return arr.filter((element) => {
        if (uniqueItems[element.place]) {
            return false;
        }

        uniqueItems[element.place] = true;

        return true;
    });
}

function getUniqueListByMap() {
    return [...new Map(arr.map(item => [item.place, item])).values()]
}

Tests:

use Map
const arr1 = getUniqueListByMap()
use Filter
const arr2 = getUniqueListByFilter()

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results

Test case name	Result
use Map
use Filter

Fastest: N/A

Slowest: N/A

Latest run results:

Run details: (Test run date: 2 years ago)

User agent: Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/121.0.0.0 Mobile Safari/537.36

Browser/OS: Chrome Mobile 121 on Android

View result in a separate tab

Test name	Executions per second
use Map	1563119.8 Ops/sec
use Filter	4669771.5 Ops/sec

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

LLMs can make mistakes. Check important info.

**Overview of the Benchmark**

The provided benchmark tests two approaches to get a unique list of properties in an object: using `Array.prototype.filter()` and using `Array.prototype.map()` with a `Map` data structure.

**Test Case 1: "use Filter"**

This test case uses the `filter()` method to remove duplicate elements from the array. The idea is to iterate over each element, check if its place property already exists in the `uniqueItems` object, and if it does, return false (indicating that the element should be skipped). If not, add the place property to the `uniqueItems` object and return true.

**Pros of using filter()**

* Efficient use of memory: The `filter()` method only iterates over each element once.
* Simple implementation: The logic is straightforward and easy to understand.

**Cons of using filter()**

* Performance overhead due to function call and conditional checks
* Potential performance impact if the array is very large

**Test Case 2: "use Map"**

This test case uses `Array.prototype.map()` in conjunction with a `Map` data structure to get a unique list of properties. The idea is to create a map where each key is a unique place property and its value is true.

**Pros of using Map()**

* Fast lookups: The `map()` method allows for fast lookups, making it suitable for large datasets.
* Efficient use of memory: The `Map` data structure only stores unique keys.

**Cons of using Map()**

* Additional memory allocation: Creating a `Map` instance requires extra memory allocation.
* Potential performance impact due to hash table operations

**Library and Special JavaScript Features**

In this benchmark, the following libraries/libraries-like-data-structures are used:

* `Array.prototype.filter()` and `Array.prototype.map()`: built-in JavaScript methods
* `Map` data structure: a native JavaScript data structure

There are no special JavaScript features or syntaxes used in this benchmark.

**Other Alternatives**

For comparison, other approaches to achieve the same goal could be:

* Using `Set` data structure instead of `Map`
* Using a different sorting algorithm (e.g., merge sort)
* Implementing a custom solution using recursive functions

However, these alternatives are not tested or implemented in this benchmark.

Related benchmarks:

Map vs filter for unique properties in object (version: 0)

Comparing performance of: use Map vs use Filter

Created: 2 years ago by: Guest

Jump to the latest result

use Map

use Filter

Suite status: <idle, ready to run>

Fastest: N/A

Slowest: N/A

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