Metadata-Version: 2.4
Name: andclust
Version: 1.0.1
Summary: ANDClust: Adaptive Neighborhood Density-Based Clustering Algorithm
Author-email: Ali Şenol <alisenol@tarsus.edu.tr>
Project-URL: Homepage, https://github.com/senolali/andclust
Project-URL: Documentation, https://github.com/senolali/andclust
Project-URL: Source, https://github.com/senolali/andclust
Project-URL: Issues, https://github.com/senolali/andclust
Project-URL: Paper, https://doi.org/10.1002/adts.202301113
Keywords: clustering,density-based,machine-learning,unsupervised-learning
Requires-Python: >=3.8
Description-Content-Type: text/markdown
Requires-Dist: numpy>=1.20
Requires-Dist: scikit-learn>=1.0
Requires-Dist: scipy>=1.8
Requires-Dist: matplotlib>=3.0

[![PyPI version](https://badge.fury.io/py/mcmstclustering.svg)](https://badge.fury.io/py/mcmstclustering)
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
[![Python 3.7+](https://img.shields.io/badge/python-3.7+-blue.svg)](https://www.python.org/downloads/)

# ANDClust
 
This package implementsthe ANDClust (Adaptive Neighborhood Distance-Based Clustering Algorithm to Cluster Varying Density and/or Neck-Typed Datasets) algorithm.

## Installation

```bash
pip install andclust
```

## Usage

```bash
from andclust import ANDClust
from sklearn.datasets import load_iris
from sklearn.preprocessing import MinMaxScaler
from sklearn.metrics.cluster import adjusted_rand_score

data=load_iris()
X,y=data['data'],data['target']

scaler = MinMaxScaler()
scaler.fit(X)
X = scaler.transform(X)

model = ANDClust(N=2,k=14,eps=0.113) # If you want to change kernel and band_with use model = ANDClust(N=2,k=14,eps=0.113,kernel='gaussian',b_width=0.025) default values for optional parameter krnl='gaussian', b_width=0.5 options for kernel are{“gaussian”, “tophat”, “epanechnikov”,
“exponential”, “linear”, “cosine”}
labels = model.fit_predict(X)

ARI=adjusted_rand_score(labels,y)
print("ARI=", ARI)

```

## Features

	- Detects arbitrary-shaped clusters due to its density-based core structure.

	- Handles varying density both between clusters and within the same cluster via a flexible neighborhood–distance mechanism.

	- Robust against outliers and noisy samples.

	- Capable of clustering high-dimensional datasets.

	- Performs well on imbalanced datasets.

	- Achieves high clustering quality across multiple evaluation metrics.

	- Effectively identifies neck-type (bottleneck-shaped) clusters.


##Citation

If you use this algorithm in research, please cite the corresponding paper.

```text
Şenol, A. (2024). ANDClust: An Adaptive Neighborhood Distance-Based Clustering Algorithm to Cluster Varying Density and/or Neck-Typed Datasets. Advanced Theory and Simulations, 7(4), 2301113.
```

#BibTeX

```text
@article{csenol2024andclust,
  title={ANDClust: An Adaptive Neighborhood Distance-Based Clustering Algorithm to Cluster Varying Density and/or Neck-Typed Datasets},
  author={{\c{S}}enol, Ali},
  journal={Advanced Theory and Simulations},
  volume={7},
  number={4},
  pages={2301113},
  year={2024},
  publisher={Wiley Online Library}
}
```
---

## LICENSE **

```text
MIT License

Copyright (c) 2025 Ali Şenol

Permission is hereby granted, free of charge, to any person obtaining a copy
...
