Showing posts with label how to scale. Show all posts
Showing posts with label how to scale. Show all posts

Thursday, June 21, 2018

How to perform best onsite on algorithm?

June 21, 2018

Introduction


It is true that I have over 300 interviews last 12 months. But in terms of real one, I understand that I only have one this year. How to push myself hard to come out a solution for those 20 minutes?

Think about scalability


It is wise to follow the hackerrank contest format. Set up time limit to 3 seconds, set up upper bound for those given variables, for example, sort a array in two dimension, I can talk about Rows <= 1000, Cols <= 1000, or given 1000,000 instead of 1000. And then I like to target the time complexity based on Rows * Cols, linear or power of 2, preprocessing time for those number, what is space minimum I need. Calculate one in terms of MB. Usually hackerrank space is limited to 300MB.

Try to guess what is limit? Ask and then figure out possible upper bound of time complexity.

I think that will be good start for an onsite and real top one software company.

I stopped playing hackerrank since November 2017. I totally forget those painful and long hours I have worked to try to gain extra few points to scale the problem. I will give one example here later on.

Get organized


Last few days I tried to get organized and checked in past contests into github, I like to review C# code I write before. And then I notice something missing in my daily practice last 8 months. I focus on soft skills, but I lose the desire to scale the algorithm problem.


Follow up 


March 28, 2019

I like to try the idea to work on algorithm problem solving 90 minutes every day. And I like to see how many problems I can solve in one week.

Another thing is to solve all 24 union join algorithm on Leetcode.com first. I need to push myself to learn and work on algorithm daily.

Another thing I can try to do is to get involved in video creating for the algorithm.

Another thing I can try to push myself is to complete another 100 algorithms before June 6, 2019. I like to celebrate one year Amazon onsite with Leetcode solved problems 470 or 500. I think that hard working is most important part of my life.