classixclustering 0.1.5

Fast and explainable clustering based on sorting

CLASSIX

Fast and explainable clustering based on sorting

CLASSIX is a fast and scalable clustering algorithm based on sorting, which results in explanable clustering. It is flexible for any shape of data with easy-tuning parameters, and share the features as follows:

• Significantly faster training speed in general cases.
• Intuitive and explainable clustering result.
• Less sensitive to data size and dimensions.
• Excellent outlier detection.
• Cython compiler Support.

The detailed document will release soon!

CLASSIX clustering consists of two phases, namely a greedy aggregation phase of the sorted data into groups of nearby data points, followed by a merging phase of groups into clusters. The algorithm is controlled by two parameters, namely the distance parameter for the aggregation and another parameter controlling the minimal cluster size. Our performance studies demonstrate that the CLASSIX achieves competing results against k-means++, DBSCAN, HDBSCAN and Quickshift++ with respect to speed and several metrics of cluster quality. In addition, CLASSIX inherent simplicity allows for the easy explainability of the clustering results. Specifically, CLASSIX runs the aggregation with user determined parameter, namely radius, to determine the initial groups by greedily selecting the heuristic centers from the whole data being sorted (the heuristic centers are formally defined as starting points in the following section). The heuristic centers, to a large extent, serve the purpose of subsampling of whole data, which are reckoned as informative as the whole data and representative with respect to the distribution of the features. Then we define density measure in the level of groups and their intersection. We define the graph on the heuristic centers by viewing the centers as vertices and an edge between each pair of vertices (weights as 1) if the density of their intersection is greater than that of any of them otherwise no edge (weith as 0). Then we find the connected components the graph as the final clusters.

Install

To install the current release via PIP use:

$ pip install ClassixClustering

Download this repository:

$ git clone https://github.com/nla-group/CLASSIX.git

Quick Start

from sklearn import datasets
from classix import CLASSIX

# Generate synthetic data
X, y = datasets.make_blobs(n_samples=1000, centers=2, n_features=2, random_state=1)

# Employ CLASSIX clustering
classix = CLASSIX(sorting='pca', radius=0.5, verbose=0)
classix.fit(X)

Get the clustering result by classix.labels_ and visualize the clustering:

plt.figure(figsize=(10,10))
plt.rcParams['axes.facecolor'] = 'white'
plt.scatter(X[:,0], X[:,1], c=classix.labels_)
plt.show()

Explainable function

CLASSIX provides convenient API for visualization of cluster assignment explanation, particularly for single object tracking and pair of object comparison. Now we demonstrate this function with example of the same data as follows:

classix.explain(plot=True)

The output clearly illustrates groups and clusters information:

A clustering of 5000 data points with 2 features has been performed. 
The radius parameter was set to 0.50 and MinPts was set to 0. 
As the provided data has been scaled by a factor of 1/6.01,
data points within a radius of R=0.50*6.01=3.01 were aggregated into groups. 
In total 7903 comparisons were required (1.58 comparisons per data point). 
This resulted in 14 groups, each uniquely associated with a starting point. 
These 14 groups were subsequently merged into 2 clusters. 
A list of all starting points is shown below.
----------------------------------------
 Group  NrPts  Cluster  Coordinates 
   0     398      0     -1.19 -1.09 
   1    1073      0     -0.65 -1.15 
   2     553      0     -1.17 -0.56 
   3     466      0     -0.67 -0.65 
   4       6      0     -0.19 -0.88 
   5       3      0     -0.72 -0.03 
   6       1      0     -0.22 -0.28 
   7     470      1       0.31 0.21 
   8     675      1       0.18 0.71 
   9     579      1       0.86 0.19 
  10     763      1       0.69 0.67 
  11       6      1       0.42 1.35 
  12       5      1       1.24 0.59 
  13       2      1        1.0 1.08 
----------------------------------------
In order to explain the clustering of individual data points, 
use .explain(ind1) or .explain(ind1, ind2) with indices of the data points. 

In the column of the simple table's, Group denotes the group label, NrPts denotes the number of data points in the associated group, Cluster is referred to as the cluster label assigned to the corresponding group, Coordinates is referred to as the coordinates of starting point associated with the group. You can easily infer the required information in this table by using following methods which we're about to demonstrate.

classix.explain(0,  plot=True)

Output: ``` The data point 0 is in group 2, which has been merged into cluster #0. ```

classix.explain(0, 2000,  plot=True)

Output: ``` The data point 0 is in group 2, which has been merged into cluster 0. The data point 2000 is in group 10, which has been merged into cluster 1. There is no path of overlapping groups between these clusters. ```

Citation

If you use CLASSIX in a scientific publication, we would appreciate your citing:

@misc{CLASSIX,
      title={Fast and explainable sorted based clustering}, 
      author={Xinye Chen and G\"{u}ttel, Stefan},
      year={2022},
      eprint={},
      archivePrefix={arXiv},
      primaryClass={}
}

Performance