Don't learn Data Structures & Algorithms for Interviews?

#neetcode #leetcode #python

[Illmare]

Sorry If I cannot come up with a project around balancing AVL trees.

[ps-dnce]

I've been coding professionally for nearly 20 years and worked in some of the most competitive companies. From my experience, this advice is beyond ridiculous. It's like telling somebody to learn geography by travelling. It simply doesn't work like that. The person giving this advice has yet to learn what they are talking about.

[ianwulf]

Honestly, i think the best advice is to do both: learn algorithms and data structures AND do personal projects (involving algorithms and data structures or not, but probably mostly not). The former is mostly to help pass technical interviews (and maybe add some tools to your programming toolbelt) while the latter will help you gain experience in building real stuff while also boosting your resume.

[musicplaylist]

What personal projects are you going to make? As a person whos been in this industry, built a bunch of projects, not once would I recommend a person to start learning through making a project. Projects get you stuck ALOT, and take ALOT of time sometimes just for little progress, sometimes you don't even learn anything after a days worth of refactoring code, finding bugs or even adjusting the specs for this "real world" half assed application. If you learn by doing projects, and avoid touching books or tutorials at some point, your learning curve is going to be a huge cliff, and let me tell yall something bout cliffs, its easy to get stuck, and even easier to fall.

[vidzpk]

I still think its much faster to learn algorithms for interviews by doing questions on said algorithms rather than coming up with novel project ideas tthat use those algorithms, the problems is that interviews requires knowing dozens of patterns and datastructures, its very difficult to find project ideas that cover a good amount of depth like that, and even if you do it'll take you months to build, while just practicing on leetcode will get you up to speed much sooner.

[derekant]

As soon as I saw this was a LinkedIn post it was all I needed to know 😂

[alluringgrace]

Linkedin is basically Facebook of professionals

[brokecoder]

He has a point.

What the linkendin guy is is saying is true with certain probability.

And highly dependent on person's objectives.

I have to add...

If you study it for interview only, you will experience the Yo-yo syndrome.

[Beaste]

In real-world applications common algorithms and data structures are mostly abstracted away and reimplementing them is just creating potential for bugs and reducing performance. The general understanding of space/time complexity and some leetcode algo patterns can help with optimizations though. Not to mention that for apps that need new kind of algorithm invention, they usually hire someone who specializes in the field as it requires way deeper knowledge and experience than any common software engineer ever gets.
If engineers were constantly working with DSA like in academics, they wouldn't need leetcode and study specially for interviews as u dont usually forget ur day to day job.

[frank]

The entire premise of this tweet is bizarre to me. You don't pick an algorithm and then build the app around it. You pick the app idea and use the patterns and algos you need to build it effectively. I'd bet most devs have never used most of these algorithms in a real world setting. Also looking at his linkedin why does he have so many jobs on there? Seems like a lot.

[priyamf]

Lol, if it goes that way, I wouldn't wake my ass up every single day at 6am in the morning to deal with problems like reversing a linked list?!! I mean all I know why I am grinding DSA IS JUST TO GET A GOOD TECH JOB..if it is improving my problem solving skills its great, but yeah. I can design and build full stack apps instead of leetcode I mean, if it wasnt a necessity. RIDICULOUS.

[DavidDLee]

Looks like this John dude was not looking for employment in a long time.
There's a vast difference between the questions you get asked in interviews and what real work entails. If you do only real work, you'll fail miserably at interviews.

[onlinealias]

Here let me go write an application implementing Floyd’s algorithm, because that’s definitely practical

[kaajjaak]

I don't think this advice is as bad as people think it is, doing this would most definitely exponentially increase your interviewing abilities, however it is definitely more of a challenge to find practical usecases for very specific algorithms than he is letting it on to be for sure

[EE]

Just your typical linkedin post. I hate when people give generic crap "advice" to look good and farm impressions instead of being blunt and practical about what will yield results.

[TRoss-rusg]

Lmao that guy’s advice is awful and doesn’t apply to entry-level applicants. And I’d be shocked if you applied trees and graphs to software as a mid-level engineer.

[Coder-jyih]

Sad thing is you cannot be a hater on linkedin

[josealejandroconcepcion]

It is a reality, you have to know data structures and design patterns for building good and well designed and reliable software components. But you have to spend time and study a lot of algorithms and patterns of problems for just come with an accept problem solutions during 45-90 mins, because its your future behind these two problems. Then you have to study some strange and wierd techniques to pass them. Even though you are not going to use most of them during your carrer. Learn coding well, learn how to create a piece of code that your temates can understand and adapt and extend in the future. For sure study hash maps, linked lists, sets, binary search, binary trees. But really you don't have to know in 30 min how to code a Kurskal algo or a segment tree to demonstrate your skill as a Software Engineer. Who the helll use this in real live projects

[rmdashrfv]

What he's saying is that if you build software that has any sort of purpose at all, it will require you to learn data structures and algorithms, and the irony that a lot of people -- a scary amount -- are missing right now is that if you simply start building useful (real-world) software, not only will you become more likely to pass interviews, but you will also become more likely to get interviews

[HARSHSHARMA-kejg]

Indian content creator made it like clearing some exam