Update app.py

app.py

@@ -5,16 +5,16 @@ import plotly.graph_objects as go
 
 CSV_URL = "https://gardenstatemls.stats.showingtime.com/infoserv/s-v1/kpou-Asg"
 
-def monte_carlo_live(T=1.0, steps=120, n_paths=1000):
+def monte_carlo_combined_plot(T=1.0, steps=120, n_paths=500):
     try:
+        # Load and clean data
         df = pd.read_csv(CSV_URL, skiprows=9)
         df = df.dropna(axis=1, how='all')
         df = df.iloc[:, :2]
         df.columns = ['Date', 'Median Sales Price']
         df['Median Sales Price'] = pd.to_numeric(df['Median Sales Price'].replace('[\$,]', '', regex=True), errors='coerce')
         df['Date'] = pd.to_datetime(df['Date'], errors='coerce')
-        df = df.dropna(subset=['Date', 'Median Sales Price'])
-        df = df.sort_values(by='Date')
+        df = df.dropna(subset=['Date', 'Median Sales Price']).sort_values(by='Date')
 
         prices = df['Median Sales Price'].values
         if len(prices) < 2:
@@ -32,69 +32,83 @@ def monte_carlo_live(T=1.0, steps=120, n_paths=1000):
             rand = np.random.normal(0, 1, n_paths)
             paths[t] = paths[t - 1] * np.exp((mu - 0.5 * sigma ** 2) * dt + sigma * np.sqrt(dt) * rand)
 
-        time = np.linspace(0, T, steps + 1)
+        forecast_dates = pd.date_range(start=df['Date'].iloc[-1], periods=steps+1, freq='MS')[1:]
 
-        forecast_fig = go.Figure()
-        for i in range(n_paths):
-            forecast_fig.add_trace(go.Scatter(x=time, y=paths[:, i], mode='lines', line=dict(width=1), showlegend=False))
+        # Combine history and forecast for unified plot
+        fig = go.Figure()
+
+        # Plot historical
+        fig.add_trace(go.Scatter(
+            x=df['Date'],
+            y=df['Median Sales Price'],
+            mode='lines+markers',
+            name='Historical',
+            line=dict(width=3)
+        ))
 
+        # Plot Monte Carlo paths
+        for i in range(n_paths):
+            full_x = pd.concat([pd.Series([df['Date'].iloc[-1]]), pd.Series(forecast_dates)])
+            full_y = np.concatenate([[S0], paths[1:, i]])
+            fig.add_trace(go.Scatter(
+                x=full_x,
+                y=full_y,
+                mode='lines',
+                line=dict(width=1),
+                showlegend=False,
+                opacity=0.3
+            ))
+
+        # Plot mean and median path
         mean_path = paths.mean(axis=1)
         median_path = np.median(paths, axis=1)
-        forecast_fig.add_trace(go.Scatter(x=time, y=mean_path, name='Mean Path', line=dict(width=3, dash='dash')))
-        forecast_fig.add_trace(go.Scatter(x=time, y=median_path, name='Median Path', line=dict(width=3, dash='dot')))
-
-        forecast_fig.update_layout(
-            title="Monte Carlo Simulation — Garden State MLS Median Sales Price Forecast",
-            xaxis_title="Years Ahead",
-            yaxis_title="Simulated Price ($)",
-            height=600
-        )
+        future_dates_full = pd.concat([pd.Series([df['Date'].iloc[-1]]), pd.Series(forecast_dates)])
 
-        historical_fig = go.Figure()
-        historical_fig.add_trace(go.Scatter(x=df['Date'], y=df['Median Sales Price'], mode='lines+markers', name='Median Sales Price'))
-        historical_fig.update_layout(
-            title="Historical Median Sales Prices (Garden State MLS)",
+        fig.add_trace(go.Scatter(x=future_dates_full, y=mean_path, name='Mean Forecast', line=dict(width=3, dash='dash')))
+        fig.add_trace(go.Scatter(x=future_dates_full, y=median_path, name='Median Forecast', line=dict(width=3, dash='dot')))
+
+        fig.update_layout(
+            title="🏡 Garden State MLS: Historical + Monte Carlo Forecast",
             xaxis_title="Date",
             yaxis_title="Median Sales Price ($)",
-            height=500
+            height=700,
+            template='plotly_white'
         )
 
-        summary = f"""**Simulation Summary**
-- Starting Price: ${S0:,.2f}
-- Simulated {n_paths} paths for {T:.1f} years
-- Annualized Drift (μ): {mu * 100:.2f}%
-- Annualized Volatility (σ): {sigma * 100:.2f}%"""
+        summary = f"""
+        **Simulation Summary**
+        - Starting Price: ${S0:,.2f}
+        - Simulated {n_paths} paths over {T:.1f} years
+        - Annual Drift (μ): {mu * 100:.2f}%
+        - Annual Volatility (σ): {sigma * 100:.2f}%
+        """
 
-        return historical_fig, forecast_fig, summary
+        return fig, summary
 
     except Exception as e:
-        err_fig = go.Figure()
-        err_fig.update_layout(title="Error", annotations=[{
+        fig = go.Figure()
+        fig.update_layout(title="Error", annotations=[{
             "text": str(e),
             "xref": "paper",
             "yref": "paper",
             "showarrow": False,
             "font": {"size": 16}
         }])
-        return err_fig, err_fig, f"Error: {e}"
+        return fig, f"Error: {e}"
 
-# Gradio UI
+# Gradio Interface
 with gr.Blocks() as demo:
-    gr.Markdown("## 🏡 Monte Carlo Simulation: Garden State MLS Median Sales Price Forecast")
+    gr.Markdown("## 📈 Garden State MLS Forecast: Historical + Monte Carlo Simulation")
 
     with gr.Row():
         years = gr.Slider(0.1, 10.0, value=1.0, step=0.1, label="Forecast Horizon (Years)")
         steps = gr.Slider(12, 240, value=120, step=12, label="Time Steps")
-        paths = gr.Slider(10, 2000, value=1000, step=10, label="Number of Simulated Paths")
+        paths = gr.Slider(10, 2000, value=500, step=10, label="Number of Simulated Paths")
 
     run_button = gr.Button("Run Simulation")
-
-    with gr.Row():
-        historical_plot = gr.Plot(label="📈 Historical Prices")
-        simulation_plot = gr.Plot(label="🔮 Forecast Simulation")
-
+    plot = gr.Plot(label="📉 Combined Historical + Forecast Plot")
     summary = gr.Markdown(label="📝 Simulation Summary")
 
-    run_button.click(fn=monte_carlo_live, inputs=[years, steps, paths], outputs=[historical_plot, simulation_plot, summary])
+    run_button.click(monte_carlo_combined_plot, inputs=[years, steps, paths], outputs=[plot, summary])
 
 demo.launch()