Advent of Code Solutions

Advent of Code is a great set of Christmas-themed coding challenges.

Usage

Requires Python >=3.10.

# Clone repo
git clone https://github.com/dshemetov/advent-of-code-solutions
cd advent-of-code-solutions

# Install (Python venv, dependencies)
make install

# Set AoC cookie in .env
make set-cookie

# Print the answer to puzzle 2021 day 2 part b
python run.py solve -y 2021 -d 2 -p b

# Print the answer to current year's day 2 (both parts)
python run.py solve -d 21

# Clear cached answer and solve again
python run.py solve -y 2021 -d 2 -c

# See help for more
python run.py --help
python run.py solve --help

Background

I first got into Advent of Code in December 2015, while spending time with the Sellin family in Los Angeles. After a few days of doing not much more than playing SSX Tricky and drinking peppermint schnapps, my brain was beginning to decompress from a tough school semester. Evin wrote functional JavaScript solutions and I learned Python. Nerding out together on a large, soft couch, I remember those days fondly.

I periodically tinker with old solutions and refactor them. The old versions can be found in the commit log, though. It's nice to look back and see yourself grow as a programmer.

• AoC 2022: 24/50 Python (with Numba and Cython).
• AoC 2021: 47/50 Python.
• AoC 2020: 32/50 Python.
• AoC 2019: 16/50 Mathematica.
• AoC 2018: 25/50 Mathematica. Lessons learned:
  • Mathematica has a some really cool builtin functions (e.g. see the three-line solution to Day 6 with DistanceMatrix and Nearest).
  • Even though Mathematica has fast built-ins, Python can be faster for simple for-loops (e.g. see Day 9 and the attached Python solution). I did not have the courage to try to implement a linked-list in Mathematica for Day 9.
  • Mathematica notebooks don't look great on GitHub.
  • Mathematica doesn't support lazy iteration natively.
  • Mathematica doesn't make it easy to make new data structures (though these are nice to have).
  • Mathematica debugging isn't easy (and it isn't easy to switch from developing code in a Notebook to Eclipse, where they maintain a debugger plugin).
• AoC 2015: 14/50 Python.