Deliberate practice techniques for software developers
A popular criticism of coding interviews is that candidates can cheat by looking up common questions and memorizing the answers. I don’t know if anyone has succeeded using that approach, but it doesn’t seem to me like it would work.
Is it possible to memorise the tricks of algorithm questions to crack the Google interview? (answer)
I’m doing most of my writing on Quora this year. For more information, see A Project for 2020.
Long ago, I wrote an article about Profiling Java Programs with VisualVM. It can be interesting to analyze code to find performance bottlenecks. But for competitive programming solutions, it’s usually a losing battle. If you choose the wrong algorithm, no amount of micro-optimization will help.
What are ways to optimize programs? My algorithms are strong enough but I am having serious difficulties in optimizations. I wonder how people in Online Judges think so creatively. (answer)
I’m doing most of my writing on Quora this year. For more information, see A Project for 2020.
People often use the term competitive programming to refer to any activity (besides homework) where programmers create solutions to made-up puzzles rather than real-world problems. For example, consider this Quora question: Which is the best online judge for competitive programming, HackerRank, HackerEarth or LeetCode? As Bohdan Pryshchenko points out in his answer, these three sites focus on interview practice, not competitive programming in the traditional sense. Their questions cover topics likely to come up in an interview. And although some of them host contests, those contests don’t consistently attract the strong contestants who regularly show up for Codeforces and Topcoder events.
When I’m writing Quora answers, I sometimes point out the difference between competitive programming and interview preparation if it’s relevant to the question. But while I’m always in favor of using terms correctly, I’m not sure how important this distinction is in the scheme of things. Consider that:
Coding interview questions require knowledge of topics like graphs, trees, and dynamic programming algorithms, all of which are used in competitive programming.
Participation in competitive programming tapers off quickly as the problems require more advanced algorithms. For example, see the bottom graph in this Quora answer. Red coders on Topcoder are a long tail at the high end of the rating distribution, while the bulk of the contestants are clustered in the large gray-green-blue area. At these levels, competitive programming questions use well-known algorithms, just like interview questions do.
Competitive programming owes most of its popularity to its association with coding interviews. If it was just about coders solving puzzles, it would be a niche activity even within the software engineering world.
As 2019 comes to a close, I’m continuing my tradition of reviewing posts from the past year. This year was divided into two halves. In the first half, I covered topics in discrete math, math for competitive programming, and math learning in general. In the second half, I changed my blog content approach. More on that later.
People debate how useful competitive programming practice is for getting better at real-world programming. But there’s one area where it definitely helps:
Why are people obsessed with competitive programming when it’s not used in real-world programming jobs? (answer)
Why is this post so short? Find the answer, and links to a blog table of contents, at A Summer 2019 Experiment.
To get better at solving coding interview problems, you have to solve a lot of coding interview problems. So a reasonable question to ask (though it’s hard to answer accurately) is:
How many LeetCode problems should I do? (answer)
Why is this post so short? Find the answer, and links to a blog table of contents, at A Summer 2019 Experiment.
This question seems basic, but it’s a core issue in math and programming practice:
What does one do when they read a problem statement, think a bit and figure they don’t know how to solve it? (answer)
Why is this post so short? Find the answer, and links to a blog table of contents, at A Summer 2019 Experiment.
In many Quora questions, people ask what’s required to start competitive programming. It’s not always clear what they’re looking for with these questions, since you don’t need to apply to be a competitive programmer. Online judges are free. Just register and try one out. Here’s a question where the writer is a bit more specific about what they’re asking:
Do I need to be good at algorithms to begin competitive programming or can I start from the bottom and get better by doing? (answer)
Why is this post so short? Find the answer, and links to a blog table of contents, at A Summer 2019 Experiment.
Some companies use a combination of automated and manual screening when deciding which candidates to invite for in-person interviews. A Quora user had a question about that:
Do companies that test their applicants using HackerRank actually look at the code or just whether they passed the test cases or not? (answer)
And here’s another question like the one from last week: How do I approach harder LeetCode problems without first looking at the solutions? (answer)
Why is this post so short? Find the answer, and links to a blog table of contents, at A Summer 2019 Experiment.
If you practice math or programming problems from a source that provides solutions, you have to decide when to peek at the solution. That might lead you to ask a question like this:
Is it still effective learning if I try medium and hard coding challenges and give up after 30 minutes and just review all the solutions if I can’t solve it myself? (answer)
Why is this post so short? Find the answer, and links to a blog table of contents, at A Summer 2019 Experiment.