Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/zhaofeng-shu33/principal_sequence_of_partition
Algorithms to compute the principal sequence of partition for Dilworth truncation function
https://github.com/zhaofeng-shu33/principal_sequence_of_partition
Last synced: about 1 month ago
JSON representation
Algorithms to compute the principal sequence of partition for Dilworth truncation function
- Host: GitHub
- URL: https://github.com/zhaofeng-shu33/principal_sequence_of_partition
- Owner: zhaofeng-shu33
- Created: 2018-10-09T12:18:58.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2023-01-23T09:33:51.000Z (almost 2 years ago)
- Last Synced: 2024-10-15T09:09:10.713Z (3 months ago)
- Language: C++
- Size: 945 KB
- Stars: 1
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Contributing: contributing.md
Awesome Lists containing this project
README
[](https://ci.appveyor.com/project/zhaofeng-shu33/principal-sequence-of-partition)
[](https://travis-ci.com/zhaofeng-shu33/principal_sequence_of_partition/)
[](https://codecov.io/gh/zhaofeng-shu33/principal_sequence_of_partition)
[](https://principal-sequence-of-partition.readthedocs.io/en/latest/?badge=latest)
# Introduction
This repository contains source code to compute the principal sequence of partition for (un)directed graph.
Currently, four methods are available:
1. Traditional Canonical Method based on Dilworth truncation and Graph Maximal Flow ([mac](https://papers.nips.cc/paper/4106-minimum-average-cost-clustering.pdf)), `dt` in our library;
2. Parametric computing scheme combined with Parametric Maximal Flow (`pdt`, our algorithm)
3. Graph Contraction combined with rapid jump (`psp_i`, our algorithm, fastest among all)
4. original parametric Dilworth truncation. ([pin](https://arxiv.org/abs/1702.00109), called `pdt_r` in our library)
All method relies on [LEMON](https://lemon.cs.elte.hu/trac/lemon) Library to compute maximum flow for graph.
# How to build
## Dependencies
* LEMON (required)
* boost-program-options is required to build the executable program `main`
* googletest (optional, used in unit-test)
You need to enable C++11 to compile the code.
### Dependency Installation Guide
* If you use CentOS/Fedora based system, you can install the pre-built binary package `liblemon` from [copr](https://copr.fedorainfracloud.org/coprs/freewind201301/liblemon/).
* If you use Ubuntu(<18.04), you can install the pre-built binary package `liblemon` from [launchpad](https://launchpad.net/~zhaofeng-shu33/+archive/ubuntu/liblemon); `liblemon` library is included in Ubuntu from 18.04, see [liblemon-dev](https://packages.ubuntu.com/bionic/liblemon-dev).
* On Windows platform, [vcpkg](https://github.com/microsoft/vcpkg) is required to fix the dependencies.
* On MacOS platform, [brew](https://brew.sh) is recommended to fix the dependencies.
## CMake
This project uses [CMake](https://cmake.org) build system. We recommend out of source build. From project root directory,
```shell
mkdir build
cd build
cmake ..
```
If any error occurs, you should fix the dependencies first.
### options
use `-DEnable_Testing=ON` to compile the test binary (requiring `googletest`).
# Reference
1. [2016] Info-Clustering: A Mathematical Theory for Data Clustering
1. [https://github.com/ktrmnm/SFM](https://github.com/ktrmnm/SFM)
1. [2010] A Faster Algorithm for Computing the Principal Sequence of Partitions of a Graph
1. [2017] Info-Clustering: An Efficient Algorithm by Network Information Flow
# Contributing
See [contributing.md](./contributing.md)
# ChangeLog
* Version 1.2: implement `DT` Class.
* Version 1.3: implement `PDT` Class
* Version 1.4: implement `PSP_I` Class.
* Version 1.5: implement `PDT_R` Class.
* Version 1.6: create uniform wrapper `PSP` Class of the above four methods