Code Editor for app.py

import streamlit as st
import numpy as np
import pandas as pd
import plotly.express as px
from sklearn.cluster import KMeans, DBSCAN
from sklearn.datasets import make_blobs
from sklearn.preprocessing import StandardScaler

# Streamlit App Title
st.title("3D Data Clustering Visualizer")

# Sidebar Controls for Data Generation
st.sidebar.header("Data Generation")
n_samples = st.sidebar.slider("Number of Samples", 100, 1000, 300, step=50)
cluster_std = st.sidebar.slider("Cluster Standard Deviation", 0.1, 3.0, 1.0, step=0.1)

# Generate synthetic data
X, _ = make_blobs(n_samples=n_samples, centers=4, cluster_std=cluster_std, random_state=42)
X = StandardScaler().fit_transform(X)  # Normalize the data

# Sidebar Controls for Clustering Algorithm Selection
st.sidebar.header("Clustering Settings")
algorithm = st.sidebar.selectbox("Choose Clustering Algorithm", ["K-means", "DBSCAN"])

# Additional Controls based on Algorithm Selection
if algorithm == "K-means":
    n_clusters = st.sidebar.slider("Number of Clusters", 2, 10, 3, step=1)
    model = KMeans(n_clusters=n_clusters, random_state=42)
    labels = model.fit_predict(X)
    centroids = model.cluster_centers_
else:
    eps = st.sidebar.slider("DBSCAN eps", 0.1, 1.0, 0.5, step=0.1)
    min_samples = st.sidebar.slider("DBSCAN min_samples", 1, 10, 5, step=1)
    model = DBSCAN(eps=eps, min_samples=min_samples)
    labels = model.fit_predict(X)
    centroids = None  # DBSCAN does not compute centroids

# Convert data to DataFrame for easier manipulation
df = pd.DataFrame(X, columns=["x", "y", "z"])
df["Cluster"] = labels.astype(str)  # Convert labels to string for better coloring in Plotly

# Plotting 3D Scatter Plot
st.header("3D Scatter Plot of Clusters")
fig = px.scatter_3d(df, x="x", y="y", z="z", color="Cluster", title=f"{algorithm} Clustering Results")

# Add cluster centroids for K-means
if algorithm == "K-means" and centroids is not None:
    centroid_df = pd.DataFrame(centroids, columns=["x", "y", "z"])
    fig.add_scatter3d(
        x=centroid_df["x"], y=centroid_df["y"], z=centroid_df["z"],
        mode="markers", marker=dict(size=10, color="black", symbol="x"),
        name="Centroids"
    )

st.plotly_chart(fig)

# Display information about the clustering
st.header("Clustering Information")
st.write("### Data Points Per Cluster")
cluster_counts = df["Cluster"].value_counts().sort_index()
st.write(cluster_counts)

if algorithm == "K-means":
    st.write("### K-means Inertia (Sum of squared distances to centroids)")
    st.write(model.inertia_)
elif algorithm == "DBSCAN":
    st.write("### DBSCAN Core Sample Count")
    core_samples = len(model.core_sample_indices_)
    st.write(f"Number of core samples: {core_samples}")