Competitive Programming

Competitive Programming

Advanced Course, 6 CP

 

Competitive Programming is the art of solving algorithmic problems and implementing solutions in a time-constrained environment. Such problems usually need insight into algorithms and efficient implementation to be solved successfully. This course will deal with general solution techniques as well as the algorithmic background knowledge to compete in programming contests. Since practice makes perfect, this course will include many tasks from previous competitions as hands-on exercises.

Aside from preparing for competitions, the goal of this lecture is to improve programming ability and apply algorithms that are usually only taught theoretically.

 

Prerequisites

You should be able to write basic programs in C++, Python or Rust. Code examples in the lecture will use C++.

Some knowledge of algorithms as taught in Grundzüge von Algorithmen und Datenstrukturen is an advantage.

You will need your own laptop to solve problems on exercise sheets and participate in the exam. If you don't have access to a laptop, please contact the lecture staff.

 

Course Preparation

If you want to prepare for the course, we recommend having a look at the book "Competitive Programming 3" which is available for registered students in our Materials section. If you already want to practice on some competitive programming tasks, you may have a look at the UVa Online Judge. The linked book recommends some problems on this website for each topic.
More problems can be found in the Codeforces problem set. Codeforces also hosts competitive programming contests about once a week.

 

Dates

Lectures: Wednesday 14:00 ct. in the Günter-Hotz lecture hall (E2.2).

 

Topics

These topics are preliminary and subject to change.

• Introduction to C++ for contests
• Basic algorithmic ideas
  • Brute-force
  • Backtracking
  • Greedy algorithms
  • Dynamic Programming
  • Divide & Conquer
• Dynamic Programming
• Graphs
  • Breadth-first search and depth-first search
  • Shortest paths
  • Topological sorting
  • Strongly connected components
  • Articulation points and bridges
  • Maximum Flow and Matching
  • Finding subgraphs
• Trees
  • Lowest common ancestor
  • Minimal spanning tree
• Mathematical Basics
• Segment trees
• String algorithms
• Parsing
• Geometric problems

 

Weekly Exercises

There will be weekly exercises. Exercises will be purely practical and consist of a set of problems (usually about 4) that have to be solved, implemented and submitted. Solutions will be graded automatically against a set of test cases. Possible verdicts are Correct, Wrong Answer (your program produced wrong output), Time Limit Exceeded (your program took too long to compute an output) or general errors such as Compilation Error or Runtime Error. You are allowed to submit solutions for exercise sheets and exams in C++, Python and Rust.

Points will be awarded iff your solution is accepted as correct by our automated judge. You will need to reach 50% of all possible points to be admitted to the exam. Internet/LLM usage for exercise sheets is strongly discouraged, but not forbidden.

 

Exams

Exams are practical contests. You are given a set of problems where you have to find, implement, and submit a solution. Exam solutions are implemented on your computer (inside of a VM provided by us with no internet connection or LLM access), submission is possible through a portal provided by us. More information on the exams and grading can be found here.