app.py

import numpy as np
import pandas as pd
from sklearn.datasets import make_classification
from sklearn.ensemble import IsolationForest
import shap
import matplotlib.pyplot as plt
from itertools import combinations
import gradio as gr

# Generate synthetic data with 20 features
np.random.seed(42)
X, _ = make_classification(
    n_samples=500,
    n_features=20,
    n_informative=10,
    n_redundant=5,
    n_clusters_per_class=1,
    random_state=42
)
outliers = np.random.uniform(low=-6, high=6, size=(50, 20))  # Add outliers
X = np.vstack([X, outliers])

# Convert to DataFrame
columns = [f"Feature{i+1}" for i in range(20)]
df = pd.DataFrame(X, columns=columns)

# Fit Isolation Forest
iso_forest = IsolationForest(
    n_estimators=100,
    max_samples=256,
    contamination=0.1,
    random_state=42
)
iso_forest.fit(df)

# Predict anomaly scores
anomaly_scores = iso_forest.decision_function(df)  # Negative values indicate anomalies
anomaly_labels = iso_forest.predict(df)  # -1 for anomaly, 1 for normal

# Add results to DataFrame
df["Anomaly_Score"] = anomaly_scores
df["Anomaly_Label"] = np.where(anomaly_labels == -1, "Anomaly", "Normal")

# SHAP Explainability
explainer = shap.Explainer(iso_forest, df[columns])
shap_values = explainer(df[columns])

# Define functions for Gradio

def get_shap_summary():
    """Generates SHAP summary plot."""
    plt.figure()
    shap.summary_plot(shap_values, df[columns], feature_names=columns, show=False)
    plt.savefig("shap_summary.png")
    return "shap_summary.png"

def get_shap_waterfall(index):
    """Generates SHAP waterfall plot for a specific data point."""
    specific_index = int(index)
    plt.figure()
    shap.waterfall_plot(
        shap.Explanation(
            values=shap_values.values[specific_index],
            base_values=shap_values.base_values[specific_index],
            data=df.iloc[specific_index],
            feature_names=columns
        )
    )
    plt.savefig("shap_waterfall.png")
    return "shap_waterfall.png"

def get_scatter_plot(feature1, feature2):
    """Generates scatter plot for two features."""
    plt.figure(figsize=(8, 6))
    plt.scatter(
        df[feature1],
        df[feature2],
        c=(df["Anomaly_Label"] == "Anomaly"),
        cmap="coolwarm",
        edgecolor="k",
        alpha=0.7
    )
    plt.title(f"Isolation Forest - {feature1} vs {feature2}")
    plt.xlabel(feature1)
    plt.ylabel(feature2)
    plt.savefig("scatter_plot.png")
    return "scatter_plot.png"

# Create Gradio interface
with gr.Blocks() as demo:
    gr.Markdown("# Isolation Forest Anomaly Detection")
    
    with gr.Tab("SHAP Summary"):
        gr.Markdown("### Global Explainability: SHAP Summary Plot")
        shap_button = gr.Button("Generate SHAP Summary Plot")
        shap_image = gr.Image()
        shap_button.click(get_shap_summary, outputs=shap_image)
    
    with gr.Tab("SHAP Waterfall"):
        gr.Markdown("### Local Explainability: SHAP Waterfall Plot")
        index_input = gr.Number(label="Data Point Index", value=0)
        shap_waterfall_button = gr.Button("Generate SHAP Waterfall Plot")
        shap_waterfall_image = gr.Image()
        shap_waterfall_button.click(get_shap_waterfall, inputs=index_input, outputs=shap_waterfall_image)
    
    with gr.Tab("Feature Scatter Plot"):
        gr.Markdown("### Feature Interaction: Scatter Plot")
        feature1_dropdown = gr.Dropdown(choices=columns, label="Feature 1")
        feature2_dropdown = gr.Dropdown(choices=columns, label="Feature 2")
        scatter_button = gr.Button("Generate Scatter Plot")
        scatter_image = gr.Image()
        scatter_button.click(get_scatter_plot, inputs=[feature1_dropdown, feature2_dropdown], outputs=scatter_image)

# Launch the Gradio app
demo.launch()