app.py

import numpy as np

from sklearn.cluster import AffinityPropagation
from sklearn import metrics
from sklearn.datasets import make_blobs

import matplotlib.pyplot as plt
import matplotlib
matplotlib.use('agg')

import gradio as gr

def generate_data(num_centers, num_samples):
    all_centers = [[1, 1], [-1, -1], [1, -1], [-1, 1]]
    centers = all_centers[:num_centers]
    
    X, labels_true = make_blobs(n_samples=num_samples, centers=centers, cluster_std=0.5, random_state=0)

    return X, labels_true


def create_plot(num_clusters, num_samples):
    X, labels_true = generate_data(num_clusters, num_samples)

    af = AffinityPropagation(preference=-50, random_state=0).fit(X)
    cluster_centers_indices = af.cluster_centers_indices_
    labels = af.labels_

    n_clusters_ = len(cluster_centers_indices)

    metrics_str = f"Estimated number of clusters: {n_clusters_}\n" 
    metrics_str += f"Homogeneity: {metrics.homogeneity_score(labels_true, labels):0.3f}\n"
    metrics_str += f"Completeness: {metrics.completeness_score(labels_true, labels):0.3f}\n"
    metrics_str += f"V-measure: {metrics.v_measure_score(labels_true, labels):0.3f}\n" 
    metrics_str += f"Adjusted Rand Index: {metrics.adjusted_rand_score(labels_true, labels):0.3f}\n" 
    metrics_str += f"Adjusted Mutual Information: {metrics.adjusted_mutual_info_score(labels_true, labels):0.3f}\n"
    metrics_str += f"Silhouette Coefficient: {metrics.silhouette_score(X, labels, metric='sqeuclidean'):0.3f}\n"
    
    fig = plt.figure(1)
    plt.clf()

    colors = plt.cycler("color", plt.cm.viridis(np.linspace(0, 1, n_clusters_)))

    for k, col in zip(range(n_clusters_), colors):
        class_members = labels == k
        cluster_center = X[cluster_centers_indices[k]]
        plt.scatter(
            X[class_members, 0], X[class_members, 1], color=col["color"], marker="."
        )
        plt.scatter(
            cluster_center[0], cluster_center[1], s=14, color=col["color"], marker="o"
        )
        for x in X[class_members]:
            plt.plot(
                [cluster_center[0], x[0]], [cluster_center[1], x[1]], color=col["color"]
            )

    plt.title("Estimated number of clusters: %d" % n_clusters_)
    plt.xlabel("x")
    plt.ylabel("y")

    return fig, metrics_str 

title = "Affinity propagation clustering algorithm"
description = """
                This demo plots clusters of a synthetic 2D dataset that contains up to 4 clusters using the affinity propagation algorithm.\
                
                
                The 2-dimensional dataset is generated around 2 to 4 predetermined cluster centers, by sampling a Gaussian distribution \
                with 0.5 standard deviation around each center. The demo uses the affinity propagation clustering algorithm to assign the data into \
                clusters. It also calculates a cluster center. \
                
                
                The figure shows a scatter plot of the data points and their connection to the respective cluster center. The demo also \
                presents several metrics based on the true and assigned labels.
            
            """
with gr.Blocks() as demo:
    gr.Markdown(f"## {title}")
    gr.Markdown(description)

    num_clusters = gr.Slider(minimum=2, maximum=4, step=1, value=2, label="Number of clusters")
    num_samples = gr.Slider(minimum=100, maximum=300, step=100, value=200, label="Number of samples")

    with gr.Row():
        plot = gr.Plot()
        text_box = gr.Textbox(label="Results")


    num_clusters.change(fn=create_plot, inputs=[num_clusters, num_samples], outputs=[plot, text_box])
    num_samples.change(fn=create_plot, inputs=[num_clusters, num_samples], outputs=[plot, text_box])

demo.launch(enable_queue=True)