re: Find Duplicates in an Array VIEW POST


While I know it's not as performant as the Map solution, would love to have seen a solution using the built-in array iterators:

const sampleData = [54,32,5,11,35,32,17,3,3,22,4,1,6,11];

const findDuplicatesUsingIterators = (sampleArr) => {
    return sampleArr.reduce((acc, curr) => {
        const duplicatesCheck = sampleArr.filter(
            (value) => curr === value;
        return duplicatesCheck.length > 1
            ? [acc, ...curr]
            : acc
    }, []);

Please excuse any typos/readability/bad code here... not a fully fleshed-out solution and I haven't run it, just wrote it real-time thinking about how to avoid explicit for loops (but there are 2 implicit loops), and I feel like it will probably be closest to, if not longer than brute force. It's a 'code golf'-ish solution that might look pretty, but actually be very inefficient.

These are the kinds of things I see (and write) a lot for readability on a known-length array, but once we get into potentially large (IE: tens of thousands+) arrays, the impact on the performance isn't well known (or at least analyzed) by the developer in a lot of cases.

Around 50k elements you will see reduce and filter (which are sometimes MORE performant than a hand-written for loop in smaller arrays) start to take longer once the function overhead takes its toll. github.com/dg92/Performance-Analys... is over two years old, and I haven't tested it myself in quite some time, but it matches my experience with it.

I LOVE these types of functions and use them all the time, but it's important to use the right tool from the toolbox.

Thanks for the article. I enjoy a good big-O analysis!


Yeap, for this first set of approaches I purposely stayed away as much as possible from the javascript built in iterators to try and get a baseline raw approach.

I was planning on a second set of tests using more JS native functions. Thanks for the code!

Code of Conduct Report abuse