This repo host our solution for the Online Qualification Round of Google Hash Code 2019.
The problem statement can be found inside of the docs folder. As a summary, we had to do the following:
Given a list of photos and the tags associated with each photo, arrange the photos into a slideshow that is as interesting as possible (the scoring section below explains what we mean by “interesting”).
Clone the repo
git clone https://github.com/serpean/ghc2019.gitMove to the project
cd ghc2019Run it
python3 main.pyShow results
Results are going to be shown after execution in this fashion (As they can be variable, this is just an example):
---------- SCORES ----------
d_pet_pictures.txt 181893
c_memorable_moments.txt 752
a_example.txt 2
b_lovely_landscapes.txt 6906
e_shiny_selfies.txt 118409
Total 307962
Note: You can add custom valid use cases to the folder input/ and they are also going to be processed.
As the time was limited and the sets were not that big, we decided to follow a greedy approach to obtain a valid solution before the time ran out. We used randomization for better sorting, so results my vary a lot between different runs. However, and for some reason, punctuation was always inside of the 300000 - 320000 range.
Out team was called Bieren lager and formed by four Computer Science students of the Universitat Politènica de València:
- Bogdan Baghiu, bogdanbaghiu - https://github.com/bogdanbaghiu
- Bullcaos, bullcaos - https://github.com/bullcaos
- Marc Solé, msolefonte - https://github.com/msolefonte
- Sergio Pérez, serpean - https://github.com/serpean
We scored 312351 points with this approach, resulting in the 140th team from Spain and the 2726 worldwide.
https://codingcompetitions.withgoogle.com/hashcode/archive/2019
Thanks to all the members of ACM UPV Chapter for allowing us spending that day together.