https://github.com/lacop/edmonds

Implementation of the Edmonds' algorithm for minimum weight perfect matching
https://github.com/lacop/edmonds

Last synced: 16 days ago
JSON representation

Implementation of the Edmonds' algorithm for minimum weight perfect matching

• Host: GitHub
• URL: https://github.com/lacop/edmonds
• Owner: lacop
• License: mit
• Created: 2015-02-08T21:24:04.000Z (almost 10 years ago)
• Default Branch: master
• Last Pushed: 2015-02-08T22:39:42.000Z (almost 10 years ago)
• Last Synced: 2023-03-10T23:08:20.558Z (almost 2 years ago)
• Language: C++
• Size: 793 KB
• Stars: 0
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
Edmonds' algorithm

==================

Implementation of Edmonds' blossom algorithm for finding minimum weight perfect matching on graphs in less than 1k SLOC.

Written as project for a class. Licensed under the MIT License.

Measurements

------------

Timed and verified with `(time ./main < input > out) && head -n 1 out`.

|Vertices | Edges | Time |

|--------:|------:|-----:|

| 1 000   | 10 495 | 500ms|

| 1 002   | 501 501| 18s  |

| 10 000  | 504 929| 4m 27s|

Testing

-------

File `test.cpp` implements simple "unit" testing with the 25 provided test cases, with the expected outcomes computed by an independent implementation:

![Testing interface](test.png)

Tested with Valgrind for memory leaks and invalid memory reads/writes.

Code coverage tested with `lcov`, implementation has 100% coverage but is reported slightly lower, see the [report](https://lacop.github.io/edmonds/).

Usage

-----

I don't recommend using this for anything serious since there are faster and more tested implementations. If you still wish to do so, refer to `main.cpp`.

To build simply run `cmake .` and then `make`. If you don't have/like cmake just build and link `main.cpp` or `test.cpp` with all other files.