• This repository provides implementations of Ex-DPC, Approx-DPC, and S-Approx-DPC.
• They are fast algorithms for density-peaks clustering (proposed in Science).
• As for the details about these algorithms, please read our SIGMOD2021 paper, Fast Density-Peaks Clustering: Multicore-based Parallelization Approach.

• spatial library
  • We used version 2.1.8.
• Boost 1.67.0
  • We have not confirmed the availability of the other version.
• The source codes of DPC algorithms have to be changed based on your paths of the above libraries.

• We prepared codes for Windows OS (Visual Studio) and Linux (Ubuntu).
• We assume low-dimensional datasets, as we use a kd-tree.
  • Implementation is based on https://github.com/gishi523/kd-tree and spatial library.

• Make a new project -> console application.
• Put our source codes into the project as with dataset, parameter, and result folders.
• Include the paths to spatial library and boost library.
• Enable OpenMP.
  • You may need /Zc:twoPhase- as an additional option at command line of C/C++ in property.
• To complie, use /Ox option.
• Set NO as SDL check.

• Ex-DPC
  • Compile: g++ -O3 Ex-DPC.cpp -o exdpc.out -fopenmp and run: ./exdpc.out.
• Approc-DPC
  • Compile: g++ -O3 main.cpp -o approxdpc.out -fopenmp and run: ./approxdpc.out.
• S-Approc-DPC
  • Compile: g++ -O3 S-Approx-DPC.cpp -o sapproxdpc.out -fopenmp and run: ./sapproxdpc.out.

• As an example, we have prepared a 2-dimensional synthetic dataset used in our paper.
• If you want to test your dataset,
  • Put the file at _dataset or dataset directory.
  • Assign a unique dataset ID.
  • Set the dimensionality at data.hpp.
  • Write codes for inputing the data file in input_data() function of file_io.hpp.
  • Add a direcotry in result and update the function compute_direcotry().
  • Compile the code and run .exe or .out file.

• Set some value in the corresponding txt file in parameter or _parameter.
• For \rho_min and \delta_min, we specify them in file_io.hpp.

• Uncomment line 211 if you need cluster labels of the exact answer.
• If you are interested in Ex-DPC+, check here.

• If you want to compute rand index, you have to run Ex-DPC and obtain the cluster labels.

If you use our implementation, please cite the following paper.

@inproceedings{amagata2021dpc,  
    title={Fast Density-Peaks Clustering: Multicore-based Parallelization Approach},  
    author={Amagata, Daichi and Hara, Takahiro},  
    booktitle={SIGMOD},  
    pages={49--61},  
    year={2021}  
}

Copyright (c) 2020 Daichi Amagata
This software is released under the MIT license.