Spencer525/IndSensor

1import streamlit as st2import pandas as pd3import numpy as np4from sklearn.cluster import KMeans, DBSCAN, AgglomerativeClustering5from sklearn.metrics import silhouette_score6from sklearn.preprocessing import StandardScaler7from statsmodels.tsa.arima.model import ARIMA8import plotly.express as px9import plotly.graph_objects as go10 11# Streamlit app title12st.title('Clustering and Time Series Analysis')13 14# Step 1: Upload CSV file15uploaded_file = st.file_uploader("Upload a CSV file", type=["csv"])16 17if uploaded_file is not None:18    data = pd.read_csv(uploaded_file)19    st.write("Dataset Preview:", data.head())20 21    # Step 2: Data Preprocessing22    # Selecting only numerical columns for clustering23    numerical_cols = data.select_dtypes(include=[np.number]).columns.tolist()24    st.write("Numerical columns for clustering:", numerical_cols)25 26    # Step 2.1: Data Standardization using StandardScaler (always applied)27    scaler = StandardScaler()28    data_scaled = scaler.fit_transform(data[numerical_cols])29    st.write("Data has been standardized using StandardScaler.")30 31    # Step 3: Clustering Algorithm Selection32    clustering_method = st.selectbox("Choose a clustering method", ["K-Means", "Hierarchical Clustering", "DBSCAN"])33 34    if clustering_method == "K-Means":35        k_range = st.slider("Select number of clusters for K-Means", min_value=2, max_value=7, value=3)36        kmeans = KMeans(n_clusters=k_range, random_state=42)37        cluster_labels = kmeans.fit_predict(data_scaled)38        silhouette_avg = silhouette_score(data_scaled, cluster_labels)39        st.write(f"K-Means Silhouette Score for {k_range} clusters: {silhouette_avg}")40 41    elif clustering_method == "Hierarchical Clustering":42        k_range = st.slider("Select number of clusters for Hierarchical Clustering", min_value=2, max_value=7, value=3)43        hierarchical = AgglomerativeClustering(n_clusters=k_range)44        cluster_labels = hierarchical.fit_predict(data_scaled)45        silhouette_avg = silhouette_score(data_scaled, cluster_labels)46        st.write(f"Hierarchical Clustering Silhouette Score for {k_range} clusters: {silhouette_avg}")47 48    elif clustering_method == "DBSCAN":49        eps_value = st.slider("Select eps value for DBSCAN", min_value=0.1, max_value=2.0, value=0.5)50        min_samples_value = st.slider("Select minimum samples for DBSCAN", min_value=1, max_value=10, value=5)51        dbscan = DBSCAN(eps=eps_value, min_samples=min_samples_value)52        cluster_labels = dbscan.fit_predict(data_scaled)53        54        # Check if DBSCAN found valid clusters55        if len(set(cluster_labels)) > 1:56            silhouette_avg = silhouette_score(data_scaled, cluster_labels)57            st.write(f"DBSCAN Silhouette Score: {silhouette_avg}")58        else:59            st.write("DBSCAN did not form valid clusters. Try adjusting eps or min_samples.")60 61    # Step 4: Visualize the clusters using Plotly62    if len(set(cluster_labels)) > 1:63        st.write("Cluster Labels:", np.unique(cluster_labels))64        65        # Create Plotly scatter plot66        fig = px.scatter(x=data_scaled[:, 0], y=data_scaled[:, 1], color=cluster_labels, title="Clustering Results", 67                         labels={'x': numerical_cols[0], 'y': numerical_cols[1]})68 69        # Update y-axis range to be from -1 to 1 with 0.2 intervals70        fig.update_layout(71            yaxis=dict(range=[-1, 1], dtick=0.2),72            xaxis_title=numerical_cols[0],73            yaxis_title=numerical_cols[1]74        )75 76        st.plotly_chart(fig)77 78    # Step 5: ARIMA Time Series Analysis79    # Checking if there are any time-related columns80    time_series_col = None81    for col in data.columns:82        if pd.api.types.is_datetime64_any_dtype(data[col]):83            time_series_col = col84            break85 86    if time_series_col:87        st.write("Time Series Analysis (ARIMA) on column:", time_series_col)88        time_series_data = data[time_series_col].dropna()89        90        # ARIMA model order91        p = st.number_input("ARIMA p value", min_value=0, max_value=5, value=1)92        d = st.number_input("ARIMA d value", min_value=0, max_value=2, value=1)93        q = st.number_input("ARIMA q value", min_value=0, max_value=5, value=1)94        95        arima_model = ARIMA(time_series_data, order=(p, d, q))96        arima_result = arima_model.fit()97        98        # Display ARIMA result summary99        st.write(arima_result.summary())100 101        # Plotting the ARIMA results102        fig = go.Figure()103        arima_result.plot_predict(dynamic=False, ax=fig.add_subplot(1, 1, 1))104        st.plotly_chart(fig)105 106    # Step 6: Create Silhouette Score Table for K-Means and Hierarchical Clustering107    st.write("### Silhouette Score Table for 2-7 Clusters")108    silhouette_scores = {'Number of Clusters': [], 'K-Means Silhouette Score': [], 'Hierarchical Silhouette Score': []}109 110    for n_clusters in range(2, 8):111        # K-Means112        kmeans = KMeans(n_clusters=n_clusters, random_state=42)113        kmeans_labels = kmeans.fit_predict(data_scaled)114        kmeans_silhouette = silhouette_score(data_scaled, kmeans_labels)115 116        # Hierarchical117        hierarchical = AgglomerativeClustering(n_clusters=n_clusters)118        hierarchical_labels = hierarchical.fit_predict(data_scaled)119        hierarchical_silhouette = silhouette_score(data_scaled, hierarchical_labels)120 121        silhouette_scores['Number of Clusters'].append(n_clusters)122        silhouette_scores['K-Means Silhouette Score'].append(kmeans_silhouette)123        silhouette_scores['Hierarchical Silhouette Score'].append(hierarchical_silhouette)124 125    silhouette_df = pd.DataFrame(silhouette_scores)126 127    # Plot the Silhouette Score Table using Plotly128    fig = go.Figure()129 130    # Plot K-Means Silhouette Scores131    fig.add_trace(go.Scatter(x=silhouette_df['Number of Clusters'], y=silhouette_df['K-Means Silhouette Score'],132                             mode='lines+markers', name='K-Means Silhouette Score'))133 134    # Plot Hierarchical Silhouette Scores135    fig.add_trace(go.Scatter(x=silhouette_df['Number of Clusters'], y=silhouette_df['Hierarchical Silhouette Score'],136                             mode='lines+markers', name='Hierarchical Silhouette Score'))137 138    # Set the y-axis range from -1 to 1 with intervals of 0.2139    fig.update_layout(140        title="Silhouette Scores for K-Means and Hierarchical Clustering",141        xaxis_title="Number of Clusters",142        yaxis_title="Silhouette Score",143        yaxis=dict(range=[-1, 1], dtick=0.2)144    )145 146    st.plotly_chart(fig)147