Update app.py

app.py

@@ -168,29 +168,31 @@ def create_threshold_analysis(df, thresholds, scope="all"):
 df = load_data()
 
 # ========== MAIN DASHBOARD ==========
-st.markdown('<h1 class="main-header">🚗 Getaround Delay Analysis</h1>', unsafe_allow_html=True)
+st.markdown('<h1 class="main-header">Getaround Delay Analysis</h1>', unsafe_allow_html=True)
 
-# Sidebar
+# Sidebar with controls
 with st.sidebar:
-    st.markdown("## 🎛️ Analysis Controls")
+    st.image("Getaround_logo.png", width=200)  
+    
+    st.markdown("## Analysis Controls")
     
     selected = option_menu(
         "Analysis Focus",
-        ["📊 Overview & Problems", "🎯 Threshold & Scope"],
-        icons=["bar-chart", "target"],
-        menu_icon="cast",
+        ["Overview & Problems", "Threshold & Scope"],
+        icons=["graph-up", "sliders"],  # More professional icons
+        menu_icon="house-gear",
         default_index=0,
     )
     
-    if selected == "🎯 Threshold & Scope":
-        st.markdown("### Controls")
-        threshold = st.slider("🕐 Threshold (minutes)", 0, 300, 90, step=30,
+    if selected == "Threshold & Scope":
+        st.markdown("### Settings")
+        threshold = st.slider("Threshold (minutes)", 0, 300, 90, step=30,
                              help="Minimum delay between consecutive rentals")
-        scope = st.selectbox("🚗 Scope", ["all", "connect"],
+        scope = st.selectbox("Implementation Scope", ["all", "connect"],
                             format_func=lambda x: "All Cars" if x == "all" else "Connect Cars Only")
     
     st.markdown("---")
-    st.markdown("### 📈 Dataset Info")
+    st.markdown("### Dataset Summary")
     st.info(f"""
     **Total Rentals:** {len(df):,}
     **Connect Rentals:** {len(df[df['checkin_type'].str.lower() == 'connect']):,}
@@ -199,11 +201,11 @@ with st.sidebar:
     """)
 
 # ========== PAGE 1: OVERVIEW & PROBLEMS ==========
-if selected == "📊 Overview & Problems":
-    st.title("📊 Understanding the Delay Problem")
+if selected == "Overview & Problems":
+    st.title("Understanding the Delay Problem")
     
     # Dataset Overview
-    st.markdown('<div class="section-header">📈 Dataset Overview</div>', unsafe_allow_html=True)
+    st.markdown('<div class="section-header">Dataset Overview</div>', unsafe_allow_html=True)
     
     col1, col2, col3, col4 = st.columns(4)
     with col1:
@@ -243,7 +245,7 @@ if selected == "📊 Overview & Problems":
         st.plotly_chart(fig_state, use_container_width=True)
     
     # Key Problem Analysis
-    st.markdown('<div class="section-header">🚨 Current Problem Scope</div>', unsafe_allow_html=True)
+    st.markdown('<div class="section-header">Current Problem Scope</div>', unsafe_allow_html=True)
     
     # Calculate current problems
     df_problems = df[df["has_previous_rental"]].copy()
@@ -290,7 +292,7 @@ if selected == "📊 Overview & Problems":
     """, unsafe_allow_html=True)
     
     # Visual analysis
-    st.markdown('<div class="section-header">📈 Delay Patterns</div>', unsafe_allow_html=True)
+    st.markdown('<div class="section-header">Delay Patterns</div>', unsafe_allow_html=True)
     
     col1, col2 = st.columns(2)
     
@@ -324,7 +326,7 @@ if selected == "📊 Overview & Problems":
         st.plotly_chart(fig_hist, use_container_width=True)
     
     # Gap analysis
-    st.markdown('<div class="section-header">📏 Time Gaps Between Rentals</div>', unsafe_allow_html=True)
+    st.markdown('<div class="section-header">Time Gaps Between Rentals</div>', unsafe_allow_html=True)
     
     gap_data = df[df["has_previous_rental"]]
     gap_filtered = gap_data[gap_data["time_delta_with_previous_rental_in_minutes"].between(0, 480)]
@@ -339,7 +341,7 @@ if selected == "📊 Overview & Problems":
     st.plotly_chart(fig_gap, use_container_width=True)
     
     # Cancellation analysis
-    st.markdown('<div class="section-header">❌ Cancellation Impact</div>', unsafe_allow_html=True)
+    st.markdown('<div class="section-header">Cancellation Impact</div>', unsafe_allow_html=True)
     
     col1, col2 = st.columns(2)
     
@@ -371,8 +373,8 @@ if selected == "📊 Overview & Problems":
             st.plotly_chart(fig_cancel, use_container_width=True)
 
 # ========== PAGE 2: THRESHOLD & SCOPE ANALYSIS ==========
-elif selected == "🎯 Threshold & Scope":
-    st.title("🎯 Threshold & Scope Decision")
+elif selected == "Threshold & Scope":
+    st.title("Threshold & Scope Decision")
     
     st.markdown("""
     <div class="insight-box">
@@ -385,7 +387,7 @@ elif selected == "🎯 Threshold & Scope":
     # Current impact
     current_metrics = calculate_metrics(df, threshold, scope)
     
-    st.markdown('<div class="section-header">📊 Impact at Current Settings</div>', unsafe_allow_html=True)
+    st.markdown('<div class="section-header">Impact at Current Settings</div>', unsafe_allow_html=True)
     
     col1, col2, col3, col4 = st.columns(4)
     with col1:
@@ -429,7 +431,7 @@ elif selected == "🎯 Threshold & Scope":
         st.caption("Formula: blocked_rentals / total_rentals * 100")
     
     # Threshold analysis
-    st.markdown('<div class="section-header">📈 Threshold Analysis</div>', unsafe_allow_html=True)
+    st.markdown('<div class="section-header">Threshold Analysis</div>', unsafe_allow_html=True)
     
     thresholds = list(range(0, 301, 30))
     threshold_data = create_threshold_analysis(df, thresholds, scope)
@@ -481,7 +483,7 @@ elif selected == "🎯 Threshold & Scope":
     st.plotly_chart(fig, use_container_width=True)
     
     # Scope comparison
-    st.markdown('<div class="section-header">🚗 Scope Comparison: All Cars vs Connect Only</div>', unsafe_allow_html=True)
+    st.markdown('<div class="section-header">Scope Comparison: All Cars vs Connect Only</div>', unsafe_allow_html=True)
     
     all_metrics = calculate_metrics(df, threshold, "all")
     connect_metrics = calculate_metrics(df, threshold, "connect")
@@ -489,7 +491,7 @@ elif selected == "🎯 Threshold & Scope":
     col1, col2 = st.columns(2)
     
     with col1:
-        st.markdown("### 🚗 All Cars")
+        st.markdown("### All Cars")
         st.markdown(f"""
         **Problems Solved:** {all_metrics['problems_solved']:,}  
         **Blocked Rentals:** {all_metrics['blocked_rentals']:,} ({all_metrics['blocked_percentage']:.1f}%)  
@@ -498,7 +500,7 @@ elif selected == "🎯 Threshold & Scope":
         """)
     
     with col2:
-        st.markdown("### 🔌 Connect Cars Only")
+        st.markdown("### Connect Cars Only")
         st.markdown(f"""
         **Problems Solved:** {connect_metrics['problems_solved']:,}  
         **Blocked Rentals:** {connect_metrics['blocked_rentals']:,} ({connect_metrics['blocked_percentage']:.1f}%)  
@@ -511,45 +513,58 @@ elif selected == "🎯 Threshold & Scope":
     connect_data = create_threshold_analysis(df, thresholds, "connect")
     
     fig_scope = go.Figure()
-    fig_scope.add_trace(go.Scatter(x=all_data["threshold"], y=all_data["problems_solved"],
-                                  mode="lines+markers", name="All Cars - Problems Solved"))
-    fig_scope.add_trace(go.Scatter(x=connect_data["threshold"], y=connect_data["problems_solved"],
-                                  mode="lines+markers", name="Connect Only - Problems Solved"))
+    
+    # Calculate percentage of problems solved for fair comparison
+    all_data["problems_solved_percent"] = (all_data["problems_solved"] / all_data["current_problems"].iloc[0] * 100).fillna(0)
+    connect_data["problems_solved_percent"] = (connect_data["problems_solved"] / connect_data["current_problems"].iloc[0] * 100).fillna(0)
+    
+    fig_scope.add_trace(go.Scatter(x=all_data["threshold"], y=all_data["problems_solved_percent"],
+                                  mode="lines+markers", name="All Cars - % Problems Solved"))
+    fig_scope.add_trace(go.Scatter(x=connect_data["threshold"], y=connect_data["problems_solved_percent"],
+                                  mode="lines+markers", name="Connect Only - % Problems Solved"))
     fig_scope.add_vline(x=threshold, line_dash="dash", line_color="red", 
                        annotation_text=f"Current: {threshold}min")
-    fig_scope.update_layout(title="Problems Solved by Scope", 
+    fig_scope.update_layout(title="Percentage of Problems Solved by Scope", 
                            xaxis_title="Threshold (minutes)", 
-                           yaxis_title="Problems Solved")
+                           yaxis_title="% of Problems Solved")
     st.plotly_chart(fig_scope, use_container_width=True)
     
     # Recommendations
-    st.markdown('<div class="section-header">💡 Recommendations</div>', unsafe_allow_html=True)
+    st.markdown('<div class="section-header">Recommendations</div>', unsafe_allow_html=True)
     
-    # Find optimal threshold
+    # Find optimal threshold - look for efficiency/revenue loss crossover point
     viable_thresholds = threshold_data[
         (threshold_data["problems_solved"] > 0) & 
         (threshold_data["threshold"] >= 30) & 
-        (threshold_data["threshold"] <= 120)  # More reasonable max
+        (threshold_data["threshold"] <= 150)
     ]
     
     if len(viable_thresholds) > 0:
-        # Find sweet spot: solve at least 80% of max problems with reasonable efficiency
-        max_problems = viable_thresholds["problems_solved"].max()
-        good_options = viable_thresholds[viable_thresholds["problems_solved"] >= max_problems * 0.8]
+        # Find crossover point where efficiency starts declining and revenue loss increases significantly
+        # Look for the "knee" where efficiency curve meets revenue loss curve
+        viable_thresholds = viable_thresholds.sort_values("threshold")
         
-        if len(good_options) > 0:
-            # Among good options, pick the one with best efficiency and lowest revenue loss
-            good_options["combined_score"] = (
-                good_options["solve_efficiency"] / 100 * 0.6 +  # 60% weight on efficiency
-                (100 - good_options["revenue_loss_percent"]) / 100 * 0.4  # 40% weight on low revenue loss
-            )
-            optimal = good_options.sort_values("combined_score", ascending=False).iloc[0]
+        # Normalize both metrics to 0-100 scale for comparison
+        viable_thresholds["efficiency_norm"] = viable_thresholds["solve_efficiency"]
+        viable_thresholds["revenue_loss_norm"] = viable_thresholds["revenue_loss_percent"]
+        
+        # Find point where efficiency is still high but revenue loss hasn't grown too much
+        # Look for threshold where efficiency > 20% and revenue loss < 8%
+        good_balance = viable_thresholds[
+            (viable_thresholds["efficiency_norm"] >= 20) & 
+            (viable_thresholds["revenue_loss_norm"] <= 8.0)
+        ]
+        
+        if len(good_balance) > 0:
+            # Among balanced options, pick the one that solves most problems
+            optimal = good_balance.sort_values("problems_solved", ascending=False).iloc[0]
         else:
-            optimal = viable_thresholds.sort_values(["problems_solved", "solve_efficiency"], ascending=[False, False]).iloc[0]
+            # Fallback: find best efficiency among viable options
+            optimal = viable_thresholds.sort_values("solve_efficiency", ascending=False).iloc[0]
         
         optimal_threshold = int(optimal["threshold"])
     else:
-        optimal_threshold = 60  # Sensible fallback
+        optimal_threshold = 90  # Sensible fallback
     
     # Scope recommendation
     if connect_metrics['solve_efficiency'] > all_metrics['solve_efficiency'] and connect_metrics['problems_solved'] >= all_metrics['problems_solved'] * 0.7:
@@ -564,17 +579,17 @@ elif selected == "🎯 Threshold & Scope":
     with col1:
         st.markdown(f"""
         <div class="recommendation-box">
-        <h3>🎯 Recommended Threshold</h3>
+        <h3>Recommended Threshold</h3>
         <h2>{optimal_threshold} minutes</h2>
         <p>Balances problem solving with availability impact</p>
-        <small>Logic: Solve ≥80% of max problems with best efficiency/revenue trade-off</small>
+        <small>Logic: Find crossover point where efficiency ≥20% and revenue loss ≤8%</small>
         </div>
         """, unsafe_allow_html=True)
     
     with col2:
         st.markdown(f"""
         <div class="recommendation-box">
-        <h3>🚗 Recommended Scope</h3>
+        <h3>Recommended Scope</h3>
         <h2>{recommended_scope}</h2>
         <p>{scope_reason}</p>
         <small>Logic: Compare efficiency and problems solved between all cars vs connect only</small>
@@ -582,7 +597,7 @@ elif selected == "🎯 Threshold & Scope":
         """, unsafe_allow_html=True)
     
     # Summary table
-    st.markdown("### 📋 Key Threshold Options")
+    st.markdown("### Key Threshold Options")
     summary_thresholds = [60, 90, 120, 150]
     summary_data = []
     
@@ -604,6 +619,6 @@ elif selected == "🎯 Threshold & Scope":
 st.markdown("---")
 st.markdown("""
 <div style='text-align: center; color: #666; padding: 1rem;'>
-<p>🚗 <strong>Getaround Delay Analysis</strong> - Supporting threshold and scope decisions</p>
+<p><strong>Getaround Delay Analysis</strong> - Supporting threshold and scope decisions</p>
 </div>
 """, unsafe_allow_html=True)