cplearn 0.1.1

Enhancing unsupervised learning with geometry-density interactions via decomposition of data into geometric core-periphery layers and subsequent clustering

cplearn

cplearn is a Python toolkit for unsupervised learning on data with underlying core–periphery-like structures.
The package includes:

• CoreSPECT – identifies most-to-least separable layers in the data w.r.t clustering, along with a clustering of most of the data
• CoreMAP – Visualization w.r.t. underlying layered structure as derived by corespect using a novel anchor-based optimization.
• Visualizer – interactive plots for visualizing core structure and subsequent layers

---

Installation

From PyPI:

pip install cplearn

---

Quickstart

import numpy as np
from sklearn.mixture import GaussianMixture

from cplearn.corespect import Corespect
from cplearn.visualizer import Visualizer as viz

# Generate synthetic 10D Gaussian Mixture data
n_samples = 500
n_components = 3
dim = 10
rng = np.random.RandomState(42)

# Define means and covariances
means = np.array([
    np.zeros(dim),
    np.ones(dim) * 2,
    np.ones(dim) * -2
])
cov = np.eye(dim) * 1.5
covariances = np.stack([cov] * n_components)

# Setup GMM manually
gmm = GaussianMixture(n_components=n_components, covariance_type='full')
gmm.weights_ = np.ones(n_components) / n_components
gmm.means_ = means
gmm.covariances_ = covariances
gmm.precisions_cholesky_ = np.linalg.cholesky(np.linalg.inv(covariances))

# Sample data
X, _ = gmm.sample(n_samples)


# Initialize and run CoreSPECT
corespect = Corespect(X)

# 1. Find dense core (e.g. 15% of data)
core = corespect.find_core(core_fraction=0.15)

# 2. Cluster core using Louvain
cluster_core = corespect.cluster_core(
    core,
    cluster_algo="louvain",
    cluster_algo_params={"ng_num": 20, "resolution": 1}
)

# 3. Propagate cluster labels outward
propagated_data = corespect.propagate_labels(
    cluster_core,
    propagate_algo="adaptive_majority", #Alternatively use "CDNN" to use the algorithm described in the CoreSPECT paper.
    propagate_algo_params={"ng_num": 20}
)

# 4. Extract layers and labels
layers, labels_for_layer = propagated_data.get_layers_and_labels()

# 5. Visualize (Visualizer internally calls CoreMAP for embedding)
mode_choice='three_steps' #use this with adaptive_majority
#You can use mode_choice='layerwise' to see a more in-depth layer-by-layer visualization
#Use layerwise if propagate_algo is "CDNN"
fig = viz(corespect,mode=mode_choice).fig
fig.show()   # or fig.write_html("corespect_viz.html")

---

References

If you use this package in your research, please cite:

• CoreSPECT
  Chandra Sekhar Mukherjee, Joonyoung Bae, and Jiapeng Zhang.
  CoreSPECT: Enhancing Clustering Algorithms via an Interplay of Density and Geometry. *link: https://arxiv.org/abs/2507.08243 *

• Recursive and adaptive majority propagation – papers coming soon

• CoreMAP – paper coming soon

Other related work

• Balanced Ranking
  Chandra Sekhar Mukherjee and Jiapeng Zhang.
  Balanced Ranking with Relative Centrality: A Multi-Core Periphery Perspective.
  ICLR 2025.

---

License

This package is licensed under the GNU General Public License v3.0 (GPLv3).
See the full license in the LICENSE file.

---