Skip to content

My solution to the reply coding challenge of 2018 - Cloud adventures

Notifications You must be signed in to change notification settings

marcocadei/reply-code-challenge18-cloud-adventures

Repository files navigation

Reply code challenge18 - Cloud adventures

My solution to the reply coding challenge of 2018 - Cloud adventures
I used Python 3 to solve the problem and numpy to help me with some vector calculus.

My scores

  • 7619
  • 1047367308
  • 252065419
  • 28624268

Scores

8th place! (out of 1207 teams)

I tried to simulate the challenge, so i gave myself just 4 hours of time: if i had partecipated i would have arrived 8th.
I'm quite happy of this result, but as a side effect the code is ugly. It should be refactored in order to be more comprehensible. In particular the choice of using tuples makes the code hard to read.
A more elegant solution would be to use named lists.

The idea

The idea i used is pretty basic: from the example in the description of the problem i noticed that the score of the considered project dropped when considering the SLA Penalty.
Therefore i ordered the projects in descending order of base penalty and bought the packages for each project in this order. In order to avoid to allocate too much packages to a single project i set a limit based on the total quantity of resources at disposal.
The idea is super easy and in fact the solution is just 106 lines long, where most of them are for the read and write functions.

About

My solution to the reply coding challenge of 2018 - Cloud adventures

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages