app.py

import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.cluster import KMeans

# 데이터 생성
np.random.seed(42)
num_samples = 30
traffic_centers = [(20, 20), (80, 80)]
nature_centers = [(0, 80), (80, 0)]
population_centers = [(0, 0), (50, 50), (100, 100)]

traffic_data = np.random.uniform(0, 100, (num_samples * len(traffic_centers), 2))
nature_data = np.random.uniform(0, 100, (num_samples * len(nature_centers), 2))
population_data = np.random.uniform(0, 100, (num_samples * len(population_centers), 2))

traffic_df = pd.DataFrame(traffic_data, columns=["x", "y"])
nature_df = pd.DataFrame(nature_data, columns=["x", "y"])
population_df = pd.DataFrame(population_data, columns=["x", "y"])

def apply_kmeans(data, k):
    kmeans = KMeans(n_clusters=k, random_state=42).fit(data)
    centroids = kmeans.cluster_centers_
    labels = kmeans.labels_
    return centroids, labels


def main():
    st.title("K-means Clustering Simulator")
    
    # Global variables declaration
    global traffic_df, nature_df, population_df

    if st.button("Initialize Datasets"):
        traffic_data = np.random.uniform(0, 100, (num_samples * len(traffic_centers), 2))
        nature_data = np.random.uniform(0, 100, (num_samples * len(nature_centers), 2))
        population_data = np.random.uniform(0, 100, (num_samples * len(population_centers), 2))

        traffic_df = pd.DataFrame(traffic_data, columns=["x", "y"])
        nature_df = pd.DataFrame(nature_data, columns=["x", "y"])
        population_df = pd.DataFrame(population_data, columns=["x", "y"])
    
    datasets = st.multiselect("Choose datasets:", ["Traffic Accessibility", "Natural Environment", "Population Density"])
    k_value = st.slider("Select k value:", 1, 10)
    
    dataset_mapping = {
        "Traffic Accessibility": (traffic_df, 'o'),
        "Natural Environment": (nature_df, 'x'),
        "Population Density": (population_df, '^')
    }
    
    combined_data = pd.concat([dataset_mapping[dataset_name][0] for dataset_name in datasets])

    centroids, labels = apply_kmeans(combined_data.values, k_value)
    
    fig, ax = plt.subplots(figsize=(8, 8))
    
    for dataset_name in datasets:
        data, marker = dataset_mapping[dataset_name]
        subset_labels = labels[:len(data)]
        ax.scatter(data['x'], data['y'], c=subset_labels, cmap='viridis', marker=marker, label=dataset_name)
        labels = labels[len(data):]
        
    ax.scatter(centroids[:, 0], centroids[:, 1], s=200, c='red', marker='X')
    
    ax.set_xlim(0, 100)
    ax.set_ylim(0, 100)
    ax.set_title(f"K-means clustering result (k={k_value})")
    ax.legend()
    st.pyplot(fig)

if __name__ == "__main__":
    main()