Metadata-Version: 2.1
Name: ImageAlgoKD
Version: 0.0.3
Summary: GPU implementation of clustering by fast search and find density peak.
Home-page: https://github.com/ZihengChen/ImageAlgorithm
Author: Ziheng Chen
Author-email: zihengchen2015@u.northwestern.edu
License: UNKNOWN
Description: # Image Algorithm for General Purpose
        
        
        Image Algorithm is a clustering algorithm based [fast search and find of density peaks](http://science.sciencemag.org/content/344/6191/1492). 
        Comparing with other popular clustering methods, such as DBSCAN,  one of the most prominent advantages of Image Algorithm is being highly parallelizable. This repository is an implementation of Image Algorithm for general purpose, supporting strong and easy GPU acceleration. 
        
        For now, the implementation includes three backends: numpy, CUDA and OpenCL.
        
        | backend | dependency | Support Platform | Support Device |
        | :---: | :---: | :---: | :---: |
        | [`numpy`](http://www.numpy.org) | None | Mac/Linux/Windows | CPU |
        | [`CUDA`](https://en.wikipedia.org/wiki/CUDA) | pycuda | Linux | Only NVIDIA GPU |
        | [`OpenCL`](https://en.wikipedia.org/wiki/OpenCL) | pyopencl | Mac | NVIDIA/AMD/Intel GPU, multi-core CPU |
        
        For all three backends, two kinds of data structure can be taken in: Flat list and kdbin. KDBin, bins in k-dimention, accommodates points inside spatial bins in k-dimention with dynamic bin distribution and flexible bin capacity. Techniqually it uses hashmap and a set of memory references to obtain nearest neighboring bins and points inside, such that query of neighborhood for each point is O(1) complexity. Performance test shows a strong acceleration in density calculation using KDBin data structure.
        
        | supported data structure |  `rho` Calculation | `rhorank` and `nh` Calculation |
        | :---: | :---: | :---: |
        | [`numpy`](http://www.numpy.org) | list/bin | list/bin |
        | [`CUDA`](https://en.wikipedia.org/wiki/CUDA) | list/bin | list |
        | [`OpenCL`](https://en.wikipedia.org/wiki/OpenCL) | list/bin | list |
        
        It has been tested that all three backends give the identical clustering results. Therefore users can feel free to choose whichever faster and easier for their purposes. Concerning speed performace, acceleration from CUDA/OpenCL may give an up to x20 speed up from CPU when dealing with more than a few thousands of data points. A preliminary speed test of three backends can be found [here](https://galleryziheng.wordpress.com/2017/12/08/gpu-acceleration-of-imaging-algorithm).
        
        
        To do list
            
        * [ ] Calculate `rhorank` in CUDA/OpenCL via 'Merge Sorted List' for parallel sorting.
        * [ ] Support `nh` search in nearest neighboring bins in CUDA/OpenCL
        
        
        ## Installation 
        
        
        ```bash
        pip install ImageAlgoKD
        ```
        
        Regarding dependency, no dependency is required for numpy backend. And it usually does a good job dealing with small dataset and needs no extra packages. However, for users wanting to use GPU acceleration with either CUDA or OpenCL backend, extra dependency is required. 
        
        ```bash
        # if want to use opencl backend
        pip install pyopencl
        # if want to use CUDA backend
        pip install pycuda
        ```
        
        ## Quick Start
        The primary usage of the module is the following
        First of all, import ImageAlgo class for K-Dimension
        ```python
        from ImageAlgoKD import *
        ```
        
        Declare an instance of ImageAlgoKD with your algorithm parameters. Then give it the input data points.
        ```python
        ia = ImageAlgoKD(MAXDISTANCE=20, KERNEL_R=1.0)
        ia.setInputsPoints(Points(np.genfromtxt("../data/basic.csv",delimiter=',')))
        ```
        
        Then run the clustering over input data points.
        ```python
        ia.run("numpy")
        # ia.run("opencl") or ia.run("cuda") if want run in parallel
        ```
        
        In the end, the clustering result can be access by
        ```python
        ia.points.clusterID
        ```
        
        ## Algorithm Parameters
        
        |     Parameters     | Comments                                                       | Default Value |
        |:------------------:|----------------------------------------------------------------|:-------------:|
        |     MAXDISTANCE    | the separation distance of the point with highest density.     |      10.0     |
        |      KERNEL_R      | 'd_c' in density calculation                                   |      1.0      |
        |    KERNEL_R_NORM   | 'd_0' in density calculation                                   |      1.0      |
        |   KERNEL_R_POWER   | 'k' in density calculation.                                    |      0.0      |
        | DECISION_RHO_KAPPA | the ratio of density threshold of seeds to the highest density |      4.0      |
        |    DECISION_NHD    | the separation threshold of seeds                              |      1.0      |
        |   CONTINUITY_NHD   | the separation threshold of continuous clusters                |      1.0      |
        
        where density is defined as
        
        <p align=center><img width="25%" src=https://github.com/ZihengChen/ImageAlgorithm/blob/master/plots/density.png /></p> 
        
        ## Examples
        
        ### I. Basic
        
        <p align=center><img width="60%" src=https://github.com/ZihengChen/ImageAlgorithm/blob/master/plots/basic.png   /></p>
        
        Perform IA clustering on 1000 toy 2D points, sampled from two Gaussian Distrituion and noise. The toy data is in `data/basic.csv`, while the corresponding jupyter notebook can be found [here](/example/example_basic.ipynb) in `examples/`.
        
        ### II. MNIST
        <p align=center> 
          <img width="40%" src=https://github.com/ZihengChen/ImageAlgorithm/blob/master/plots/mnist_decision.png /> 
          <img width="45%" src=https://github.com/ZihengChen/ImageAlgorithm/blob/master/plots/mnist.png />
        </p> 
        
        Perform IA clustering on 1000 MNIST 28x28 dimension points. The MNIST data is in `data/mnist.csv`, while the corresponding jupyter notebook can be found [here](/example/example_mnist.ipynb) in `examples/`.
        
        ### III. HGCal
        
        
        <p align=center><img width="50%" src=https://github.com/ZihengChen/ImageAlgorithm/blob/master/plots/hgcal.png /></p> 
        
        This is an event of 10 Pions with 300 GeV energy in CMS HGCal. A 3D interactive visualization can be found [here](https://plot.ly/~zihengchen/61/#/). In addition, for event with pile up, [here](https://plot.ly/%7Ezihengchen/18/#/) is an 300GeV pion with PU200 event. A PU200 event typically includes about 200k HGVCal reconstructed detector hits, which is input into IA clustering
        
        
        
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Description-Content-Type: text/markdown
