This is the unnoficial repo for calculating running times of helmes challenge solutions. Benchmark tries to not allow app use a cached version of file, as it was explicitely stated that file should be referenced from disk. Right now it does so by keeping copies of the same dictionary files. Obviously timing will be differt on different machines, so you can run this on yours (be careful with untrusted solutions).

Open an issue if you want to add your solution in results.

The result should be an executable application, that finds all anagrams for provided word from a provided dictionary as fast as possible. The application is responsible for calculating the execution time and returning it in microseconds (μs). We will be testing your code against the dictionary found here: (All files are the same in that directory)

git clone https://github.com/scientist1642/anagram_benchmarks --recursive

python3 benchmark.py [repo_name | all] python3 benchmark.py all - benchmark all solutions - takes a long time python3 benchmark.py username/reponame - benchmark a specific solution.

To generate markdown table:

cat benchmark_results.csv | python3 tools/mdtable.py -separator:,

Update repos.txt and git submodule add https://github.com/adamsont/helmes_challenge.git solutions/adamsont/helmes_challenge

Drop cache before each run with echo 3 > /proc/sys/vm/drop_cachesa

Done on ubuntu 18.04 64bit virtualbox vm 2 cpu / Macos 2014 2ghz

ubuntu 18.04 ryzen 1700x 8 core

i7-7700HQ

i7-6700HQ