The exact answers to leetcode questions are probably included many times in the Copilot dataset, in many programming languages. Fresh Advent of code problems are much more interesting.
There's also a wide variety of Leetcode problems, with some (data transformation and validation problems in particular) being more resembling of real life coding than others (e.g. dynamic programming problems).
There seem to still be interviewers out there who focus on puzzle-like problems.
Not sure what you meant. Leetcode type things solves a problem of past. Future will be copilots doing lot of grunt coding work and human value add would be instructing copilots the right way, bringing broader contextual information. Solving binary search tree problems during interviews will eventually go away (once old guards go away)
I'm pretty surprised that LeetCode hasn't been mentioned. There are a lot of low quality questions on it but if you sort by "Top 100", you'll find some pretty interesting problems.
The point of a leetcode style question is to test someones' ability to explain and adapt a tricky problem on the fly. Given most of a software engineer's day to day is about communication and adapting existing solutions to related problems, leetcode problems are a very powerful tool for analyzing candidates.
I try to do a few LeetCode problems a day as my own take on the “coding kata” idea. There is really a pretty small set of problems and subproblems that crop up over and over again in LC problems, so stuff like building a tree or graph traversal or performing a string or array transformation become very well worn and familiar tools.
leetcode is similar to the hackerrank website, it's a set of about 700 programming problems where one can enter a solution in almost any programming language applicable (there are some SQL queries). There is also a discussion forum for each question. Occasionally there is more writing by the problem creator with alternate solutions listed. I've found a few companies use the hackerrank for initial programming tests, I kind of dislike hackerrank because alot of the hackerrank tests rely on stdio input equivalent in each language which in some languages I never use in real life, i.e. I never use the Java command line prompt except for fooling around on hackerrank and some of the questions on hackerrank are ridiculously obtuse - I've skipped to the discussion and find that is the number one upvoted type of comment more than once.
I literally was asked two leetcode questions verbatim when interviewing for a data science position at TikTok a few months ago. Dynamic programming (I won't mention which question) and then one regarding binary trees.
LeetCode is nice because it shows you other people's solutions when you correctly submit. Super useful for seeing code with faster execution times, I've learned some tricks from that.
The predecessors to leetcode questions were being asked, for example, in PhD defense dissertations as well as some that were highly specific to the company in questions (sort 2MB of data 1MB RAM). Many of the problems are basically late-undergrad, early-graduate CS student problems.
reply