Edit Distance

This repository contains code for our paper "Efficient Parallel Output-Sensitive Edit Distance".

Requirements

• CMake >= 3.15
• g++ or clang with C++17 features support (tested with g++ 12.1.1 and clang 14.0.6) on Linux machines.
• We use ParlayLib to support fork-join parallelism and some parallel primitives. It is provided as a submodule in our repository.

Getting Code

git clone --recurse-submodules https://github.com/ucrparlay/Edit-Distance.git

Compilation

mkdir build && cd build
cmake ..
make

Running Code

For synthetic dataset:

./test_framework <id> <n> <k> <sigma> <rounds>

• id: id of the algorithm
  • id=-1 all of our algorithms [BFS-Hash, BFS-H-Hash, BFS-SA, DaC-SP]
  • id=0: BFS-Hash
  • id=1: BFS-H-Hash
  • id=2: BFS-SA
  • id=3: DaC-SP
  • id=4: Sequential DP
  • id=5: ParlayLib
• n: estimated length of strings
• k: estimated number of edits
• sigma: alphabet size
• rounds: number of rounds

For example, to run all our algorithms on $n=10^9$ and $k=10$:

./test_framework -1 1000000000 10 256

For real-world dataset:

./test_framework_real -i <id> -f1 <path to file1> -f2 <path to file2> 

Reference

Xiangyun Ding, Xiaojun Dong, Yan Gu, Youzhe Liu and Yihan Sun. Efficient Parallel Output-Sensitive Edit Distance. To appear at European Symposium on Algorithms (ESA), 2023

If you use our code, please cite our paper:

@inproceedings{ding2023efficient,
  author    = {Ding, Xiangyun and Dong, Xiaojun and Gu, Yan and Sun, Yihan and Liu, Youzhe},
  title     = {Efficient Parallel Output-Sensitive Edit Distance},
  booktitle = {European Symposium on Algorithms (ESA)},
  year      = {2023}
}