Update app.py

app.py

@@ -7,7 +7,7 @@ from plotly.subplots import make_subplots
 import folium
 from streamlit_folium import st_folium
 from sklearn.linear_model import LinearRegression
-
+import os 
 # ================= CONFIG =================
 st.set_page_config(
     page_title="Michelin Mining Tyre Analytics",
@@ -243,6 +243,68 @@ def load_data():
     return df
 
 df = load_data()
+# @st.cache_data
+# def load_data():
+#     try:
+#         # Load main data
+#         df = pd.read_excel("df_final.xlsx", sheet_name="Sheet1")
+#         # Load health index data
+#         hi_data = pd.read_excel("hi_final.xlsx")
+#     except FileNotFoundError as e:
+#         st.error(f"❌ File not found: `{e.filename}`")
+#         st.stop()
+#     except Exception as e:
+#         st.error(f"❌ Error loading data: {e}")
+#         st.stop()
+    
+#     # === Proses df_final.xlsx ===
+#     # Fix encoding
+#     df.columns = df.columns.str.replace("Â", "")
+#     for col in df.select_dtypes(include='object').columns:
+#         df[col] = df[col].astype(str).str.replace("Â", "")
+    
+#     # Parse datetime
+#     df['Time'] = pd.to_datetime(df['Time'], errors='coerce')
+#     df = df.dropna(subset=['Time']).copy()
+#     df['hour'] = df['Time'].dt.hour
+    
+#     # Alarm flag
+#     df['is_alarm'] = (~df['Alarm Status'].fillna('').str.contains('No Alarm', case=False)).astype(int)
+    
+#     # Dynamic risk score
+#     p = df['Pressure (psi)']
+#     p_red_high = df['Red High Press (psi)']
+#     p_amber_high = df['Amber High Press (psi)']
+#     t = df['Temperature (°C)']
+#     t_red = df['Absolute Red Temp (°C)']
+#     t_amber = df['Absolute Amber Temp (°C)']
+    
+#     # Avoid division by zero
+#     p_denom = (p_red_high - p_amber_high).replace(0, np.nan)
+#     p_norm = np.clip((p - p_amber_high) / p_denom, 0, 1).fillna(0)
+#     t_denom = (t_red - t_amber).replace(0, np.nan)
+#     t_norm = np.clip((t - t_amber) / t_denom, 0, 1).fillna(0)
+#     df['risk_score'] = 0.6 * p_norm + 0.4 * t_norm
+
+#     def get_risk_label(score):
+#         if score >= 0.8: return 'Very High Risk'
+#         elif score >= 0.6: return 'High Risk'
+#         elif score >= 0.3: return 'Moderate Risk'
+#         else: return 'Slight Risk'
+#     df['Risk Level'] = df['risk_score'].apply(get_risk_label)
+    
+#     # Position Group
+#     df['Position Group'] = df['Position'].map({1: 'Front', 2: 'Front', 3: 'Rear', 4: 'Rear'}).fillna('Other')
+
+#     return df, hi_data  # Kembalikan 2 data
+# # Load both datasets
+# df, hi_data = load_data()
+# # Optional: Info ringkas di sidebar (opsional)
+# with st.sidebar:
+#     st.markdown("### 📊 Dataset Overview")
+#     st.metric("Total Records", f"{len(df):,}")
+#     st.metric("Date Range", f"{df['Time'].min().date()} → {df['Time'].max().date()}")
+#     st.metric("Alarms (Red/Amber)", f"{df['is_alarm'].sum():,} ({df['is_alarm'].mean():.1%})")
 
 # ================= HEADER =================
 st.markdown("""
@@ -400,7 +462,7 @@ st.markdown('<h3 class="objective-title">OBJECTIVE 1: Pressure & Temperature Tre
 col2, col1 = st.columns(2)
 
 # Define consistent color mapping
-color_map = {1: '#003DA5', 2: '#7FA6E8', 3: '#FFB300', 4: '#FFE082'}
+color_map = {1: '#003DA5', 2: '#7FA6E8', 3: '#3F7F73', 4: '#8EC3B7'}
 category_order = [1, 2, 3, 4]
 
 with col1:
@@ -551,7 +613,7 @@ def create_radial_chart(pos_data, title, shift_hours, shift_type):
         r=hourly_amber.values,
         theta=theta,
         name='Amber',
-        marker_color='#FFC107',  # Kuning
+        marker_color='#FFBF00',  # Kuning
         opacity=0.8,
         hovertemplate='<b>Hour:</b> %{customdata}:00<br><b>Count:</b> %{r}<br><b>Status:</b> Amber<extra></extra>',
         customdata=shift_hours
@@ -679,88 +741,188 @@ with col8:
     else:
         st.warning("No data for Position 4 (18:00–06:00)")
 
+def generate_objective2_insights(df: pd.DataFrame) -> str:
+    """
+    Generate dynamic, line-by-line insight for Objective 2.
+    Input: filtered DataFrame (e.g., by site/year/hour — already applied upstream)
+    Output: formatted string with <br> for newline (left-aligned, clean)
+    """
+    # Filter hanya Red & Amber alarms
+    alarm_df = df[df['alarm_severity'].isin(['Red', 'Amber'])].copy()
+    
+    if alarm_df.empty:
+        return "• No Red or Amber alarms in selected period."
+
+    # === 1. Total counts ===
+    total_red = alarm_df[alarm_df['alarm_severity'] == 'Red'].shape[0]
+    total_amber = alarm_df[alarm_df['alarm_severity'] == 'Amber'].shape[0]
+    total_combined = total_red + total_amber
+
+    # === 2. Dominant period (custom bins) ===
+    def get_period(hour):
+        if 12 <= hour < 18:
+            return '12:00–18:00'
+        elif 18 <= hour < 24 or hour == 0:  # include 00:00 as part of night
+            return '18:00–00:00'
+        else:
+            return 'Other'
+    
+    alarm_df['period'] = alarm_df['hour'].apply(get_period)
+    
+    period_counts = alarm_df['period'].value_counts()
+    dominant_period = period_counts.idxmax()
+    dominant_pct = period_counts.max() / total_combined * 100
+    
+    # Second-dominant
+    if len(period_counts) > 1:
+        second_period = period_counts.index[1]
+        second_pct = period_counts.iloc[1] / total_combined * 100
+    else:
+        second_period = '—'
+        second_pct = 0.0
+
+    # === 3. Peak per Position + Period + Severity ===
+    # We'll find peak hour within each (Position, period, severity) group
+    peaks = []
+    for pos in [1, 2, 3, 4]:
+        for period in ['12:00–18:00', '18:00–00:00']:
+            for sev in ['Red', 'Amber']:
+                subset = alarm_df[
+                    (alarm_df['Position'] == pos) &
+                    (alarm_df['period'] == period) &
+                    (alarm_df['alarm_severity'] == sev)
+                ]
+                if not subset.empty:
+                    peak_hour = subset['hour'].mode().iloc[0]  # most frequent hour
+                    peak_count = (subset['hour'] == peak_hour).sum()
+                    # Format hour to 00:00 (0 → 00:00, 23 → 23:00, but 0 in 18–00 group = 00:00)
+                    display_hour = f"{int(peak_hour):02d}:00"
+                    peaks.append((pos, period, sev, display_hour, peak_count))
+
+    # Build insight lines
+    lines = [
+        f"• Dominant period for Red/Amber: {dominant_period} — {dominant_pct:.2f}%",
+        f"• Second-dominant period: {second_period} — {second_pct:.2f}%"
+    ]
+
+    # Add peak lines (only top patterns — limit to meaningful ones)
+    for pos, period, sev, hr, cnt in sorted(peaks, key=lambda x: (-x[4], x[0], x[1])):
+        if cnt >= 10:  # only show peaks with ≥10 occurrences (avoid noise)
+            lines.append(f"• Position {pos}, {period}: Peak {sev} alarm at {hr} ({cnt:,} occurrences)")
+
+    return "<br>".join(lines)
+
+
+# === RENDER INSIGHT BOX ===
 # =============== INSIGHT 2 (Ringkas & Fokus ke Red & Amber) ===============
 if alarm_data.empty:
-    insight_text = "• No data available for analysis."
+    insight_lines = ["• No alarm data available for Red or Amber analysis."]
 else:
-    # Insight tetap sama
-    alarm_hours = alarm_data['hour']
-
-    def hour_to_band(h):
-        if 0 <= h < 6:   return "00:00–06:00 (Night)"
-        if 6 <= h < 12:  return "06:00–12:00 (Morning)"
-        if 12 <= h < 18: return "12:00–18:00 (Afternoon)"
-        return "18:00–00:00 (Evening)"
-
-    alarm_hours_df = pd.DataFrame({'hour': alarm_hours})
-    alarm_hours_df['band'] = alarm_hours_df['hour'].apply(hour_to_band)
-    band_counts = alarm_hours_df['band'].value_counts().sort_index()
-
-    top_bands = band_counts.nlargest(2)
-    dominant_band = top_bands.index[0] if len(top_bands) > 0 else "N/A"
-    second_dominant_band = top_bands.index[1] if len(top_bands) > 1 else "N/A"
-
-    dominant_pct = (top_bands.iloc[0] / band_counts.sum() * 100) if len(top_bands) > 0 else 0
-    second_pct = (top_bands.iloc[1] / band_counts.sum() * 100) if len(top_bands) > 1 else 0
-
-    # Hitung jumlah masing-masing jenis alarm
-    normal_alarms = alarm_data[alarm_data['Alarm Status'] == 'No Alarm'].shape[0]
-    red_alarms = alarm_data[alarm_data['Alarm Status'].str.contains('Red', na=False)].shape[0]
-    amber_alarms = alarm_data[alarm_data['Alarm Status'].str.contains('Amber', na=False)].shape[0]
-
-    # Insight Spesifik Per Position dan Shift
-    insight_lines = [
-        f""
-    ]
+    # Filter hanya Red dan Amber (case-insensitive, robust terhadap NaN)
+    red_amber_data = alarm_data[
+        alarm_data['Alarm Status'].str.contains(r'\b(Red|Amber)\b', case=False, na=False)
+    ].copy()
+
+    if red_amber_data.empty:
+        insight_lines = ["• No Red or Amber alarms detected in the dataset."]
+    else:
+        # Hitung jumlah
+        red_alarms = red_amber_data['Alarm Status'].str.contains('Red', case=False).sum()
+        amber_alarms = red_amber_data['Alarm Status'].str.contains('Amber', case=False).sum()
+        total_red_amber = len(red_amber_data)
+
+        # Kelompok waktu dominan: hanya 2 band utama (sesuai preferensi)
+        def hour_to_band(h):
+            if 12 <= h < 18:
+                return "12:00–18:00 (Afternoon)"
+            elif (18 <= h <= 23) or (0 <= h < 6):
+                return "18:00–00:00 (Evening/Night)"
+            else:
+                return "06:00–12:00 (Morning)"  # fallback, jarang muncul
+
+        red_amber_data['time_band'] = red_amber_data['hour'].apply(hour_to_band)
+        band_counts = red_amber_data['time_band'].value_counts()
+        top2 = band_counts.nlargest(2)
+
+        # Ambil dominan & kedua dominan (jika ada)
+        dominant_band = top2.index[0] if len(top2) > 0 else "—"
+        second_band = top2.index[1] if len(top2) > 1 else "—"
+        dom_pct = (top2.iloc[0] / total_red_amber * 100) if len(top2) > 0 else 0
+        sec_pct = (top2.iloc[1] / total_red_amber * 100) if len(top2) > 1 else 0
+
+        # Insight utama
+        insight_lines = [
+            # f"• Total Red alarms: {red_alarms}, Amber alarms: {amber_alarms} (combined: {total_red_amber})",
+            # f"• Dominant time band: {dominant_band} ({dom_pct:.1f}%)",
+            # f"• Second-dominant time band: {second_band} ({sec_pct:.1f}%)"
+        ]
+
+        # Helper: peak hour (Red/Amber) dalam subset
+        def peak_hour_in(data, alarm_type):
+            subset = data[data['Alarm Status'].str.contains(alarm_type, case=False, na=False)]
+            if not subset.empty:
+                counts = subset['hour'].value_counts()
+                h = int(counts.idxmax())
+                c = int(counts.max())
+                return h, c
+            return None, 0
+
+        # Loop Position 1–4 (sesuai preferensi: breakdown per position)
+        for pos in [1, 2, 3, 4]:
+            # Band 1: 12:00–18:00
+            band1 = red_amber_data[
+                (red_amber_data['Position'] == pos) &
+                (red_amber_data['hour'].between(12, 17))
+            ]
+            # Band 2: 18:00–00:00
+            band2 = red_amber_data[
+                (red_amber_data['Position'] == pos) &
+                ((red_amber_data['hour'] >= 18) | (red_amber_data['hour'] <= 5))
+            ]
+
+            # Band 1
+            if not band1.empty:
+                h_r, c_r = peak_hour_in(band1, 'Red')
+                if c_r > 0:
+                    insight_lines.append(f"• Position {pos}, 12:00–18:00: Peak Red alarm at {h_r:02d}:00 ({c_r} occurrences).")
+                h_a, c_a = peak_hour_in(band1, 'Amber')
+                if c_a > 0:
+                    insight_lines.append(f"• Position {pos}, 12:00–18:00: Peak Amber alarm at {h_a:02d}:00 ({c_a} occurrences).")
+
+            # Band 2
+            if not band2.empty:
+                h_r, c_r = peak_hour_in(band2, 'Red')
+                if c_r > 0:
+                    insight_lines.append(f"• Position {pos}, 18:00–00:00: Peak Red alarm at {h_r:02d}:00 ({c_r} occurrences).")
+                h_a, c_a = peak_hour_in(band2, 'Amber')
+                if c_a > 0:
+                    insight_lines.append(f"• Position {pos}, 18:00–00:00: Peak Amber alarm at {h_a:02d}:00 ({c_a} occurrences).")
+
+# =============== DISPLAY (NEW LINE PER BULLET) ===============
+insight_text = "\n".join(insight_lines)
 
-    # Fungsi helper untuk mencari jam dengan count maksimum untuk alarm Red/Amber
-    def get_max_alarm_hour(pos_data, alarm_type):
-        filtered_data = pos_data[pos_data['Alarm Status'].str.contains(alarm_type, na=False)]
-        if not filtered_data.empty:
-            hourly_counts = filtered_data.groupby('hour').size()
-            if not hourly_counts.empty:
-                max_hour = hourly_counts.idxmax()
-                max_count = hourly_counts.max()
-                return max_hour, max_count
-        return None, 0
-
-    # Loop untuk semua Position (1, 2, 3, 4)
-    for pos in [1, 2, 3, 4]:
-        # Shift Pagi (06:00–17:59)
-        pos_pagi = alarm_data[(alarm_data['Position'] == pos) & (alarm_data['hour'].between(6, 17, inclusive='both'))]
-        if not pos_pagi.empty:
-            # Cari jam dengan alarm Red maksimum
-            red_hour, red_count = get_max_alarm_hour(pos_pagi, 'Red')
-            if red_hour is not None:
-                insight_lines.append(f"Position {pos} Shift 1 (06:00–18:00): Peak Red alarm at {red_hour:02d}:00 ({red_count} alarms).")
-            # Cari jam dengan alarm Amber maksimum
-            amber_hour, amber_count = get_max_alarm_hour(pos_pagi, 'Amber')
-            if amber_hour is not None:
-                insight_lines.append(f" Position {pos} Shift 1 (06:00–18:00): Peak Amber alarm at {amber_hour:02d}:00 ({amber_count} alarms).")
-
-        # Shift Sore (18:00–05:59)
-        pos_sore = alarm_data[(alarm_data['Position'] == pos) & ((alarm_data['hour'] >= 18) | (alarm_data['hour'] <= 5))]
-        if not pos_sore.empty:
-            # Cari jam dengan alarm Red maksimum
-            red_hour, red_count = get_max_alarm_hour(pos_sore, 'Red')
-            if red_hour is not None:
-                insight_lines.append(f" Position {pos} Shift 2 (18:00–06:00): Peak Red alarm at {red_hour:02d}:00 ({red_count} alarms).")
-            # Cari jam dengan alarm Amber maksimum
-            amber_hour, amber_count = get_max_alarm_hour(pos_sore, 'Amber')
-            if amber_hour is not None:
-                insight_lines.append(f" Position {pos} Shift 2 (18:00–06:00): Peak Amber alarm at {amber_hour:02d}:00 ({amber_count} alarms).")
-
-    insight_text = "\n".join(insight_lines)
-
-# =============== DISPLAY INSIGHT ===============
 st.markdown(f"""
 <div class="insight-box">
-    <div class="content">
+    <div class="content" style="
+        text-align: left;
+        white-space: pre-line;
+        font-family: 'Segoe UI', sans-serif;
+        line-height: 1.6;
+    ">
 {insight_text}
     </div>
+    <div style="
+        font-size: 0.85em;
+        color: #6C757D;
+        margin-top: 16px;
+        text-align: right;
+        font-weight: 500;
+    ">
+    </div>
 </div>
 """, unsafe_allow_html=True)
 #### OBJECTICVE 3
+# ================= OBJECTIVE 3 =================
 st.markdown('<h3 class="objective-title">OBJECTIVE 3: Correlation — How Does Heat Influence Pressure and Which Tyres Trigger Red Alarms?</h3>', unsafe_allow_html=True)
 
 # Prepare data
@@ -769,341 +931,284 @@ rear_df = dff[dff['Position'].isin([3, 4])].copy()
 
 col1, col2 = st.columns(2)
 
-# =============== COL 1: Front — Temperature → Pressure (Scatter + Regression Area) ===============
+# =============== COL 1: Front — Temp → Pressure ===============
 with col1:
     st.markdown('<h5 style="text-align:center; margin-top: 0;">Front Tyres: Temperature → Pressure</h5>', unsafe_allow_html=True)
     
     if not front_df.empty:
-        # Tambahkan kategori alarm status
         front_df['Category'] = front_df.apply(
-            lambda row: f"Normal Front Tyre" if row['Alarm Status'] == 'No Alarm'
-            else f"Amber Pressure Front Tyre" if 'Amber' in row['Alarm Status']
-            else f"Red Pressure Front Tyre", axis=1
+            lambda row: "Normal Front Tyre" if row['Alarm Status'] == 'No Alarm'
+            else "Amber Pressure Front Tyre" if 'Amber' in str(row['Alarm Status'])
+            else "Red Pressure Front Tyre", axis=1
         )
         categories = ["Normal Front Tyre", "Amber Pressure Front Tyre", "Red Pressure Front Tyre"]
         front_df['Category'] = pd.Categorical(front_df['Category'], categories=categories, ordered=True)
 
-        # Filter valid data
-        valid_data = front_df.dropna(subset=['Temperature (°C)', 'Pressure (psi)'])
-        if len(valid_data) > 1:
-            X = valid_data[['Temperature (°C)']]
-            y = valid_data['Pressure (psi)']
+        valid = front_df.dropna(subset=['Temperature (°C)', 'Pressure (psi)'])
+        if len(valid) > 1:
+            X = valid[['Temperature (°C)']].values
+            y = valid['Pressure (psi)'].values
             model = LinearRegression().fit(X, y)
             x_line = np.linspace(X.min(), X.max(), 100).reshape(-1, 1)
             y_line = model.predict(x_line)
-            corr = np.corrcoef(valid_data['Temperature (°C)'], valid_data['Pressure (psi)'])[0, 1]
+            corr = np.corrcoef(valid['Temperature (°C)'], valid['Pressure (psi)'])[0, 1]
 
-            fig1 = px.scatter(
-                valid_data,
+            fig = px.scatter(
+                valid,
                 x='Temperature (°C)',
                 y='Pressure (psi)',
                 color='Category',
                 color_discrete_map={
-                    "Normal Front Tyre": "#2E7D32",         # Hijau
-                    "Amber Pressure Front Tyre": "#FFC107", # Kuning
-                    "Red Pressure Front Tyre": "#D32F2F"    # Merah
+                    "Normal Front Tyre": "#2E7D32",
+                    "Amber Pressure Front Tyre": "#FFBF00",
+                    "Red Pressure Front Tyre": "#D32F2F"
                 },
                 category_orders={'Category': categories},
-                template="plotly_white",
-                labels={'Temperature (°C)': 'Temperature (°C)', 'Pressure (psi)': 'Pressure (psi)'}
+                template="plotly_white"
             )
-
-            fig1.update_traces(
-                hovertemplate="<b>%{marker.color}</b><br>Temp: %{x:.1f}°C<br>Pressure: %{y:.1f} psi<extra></extra>",
+            fig.update_traces(
+                hovertemplate="<b>%{customdata[0]}</b><br>Temp: %{x:.1f}°C<br>Pressure: %{y:.1f} psi<extra></extra>",
+                customdata=valid[['Category']].values,
                 marker=dict(size=6)
             )
-
-            fig1.add_trace(go.Scatter(
+            fig.add_trace(go.Scatter(
                 x=x_line.flatten(), y=y_line,
                 mode='lines', name='Trend Line',
                 line=dict(color='#1976D2', dash='dot', width=2)
             ))
 
-            # Tambahkan area confidence interval (soft background)
+            # Confidence band
             y_pred = model.predict(X)
-            residuals = y - y_pred
-            mse = np.mean(residuals**2)
-            std_error = np.sqrt(mse)
-            y_upper = y_line + 1.96 * std_error
-            y_lower = y_line - 1.96 * std_error
-
-            fig1.add_trace(go.Scatter(
+            std_err = np.std(y - y_pred)
+            y_upper = y_line + 1.96 * std_err
+            y_lower = y_line - 1.96 * std_err
+            fig.add_trace(go.Scatter(
                 x=np.concatenate([x_line.flatten(), x_line.flatten()[::-1]]),
                 y=np.concatenate([y_upper, y_lower[::-1]]),
                 fill='toself',
-                fillcolor='rgba(211, 47, 47, 0.1)',  # Merah transparan
+                fillcolor='rgba(211, 47, 47, 0.1)',
                 line=dict(color='rgba(255,255,255,0)'),
-                showlegend=False,
-                name='Confidence Interval'
+                showlegend=False
             ))
 
-            fig1.update_layout(
+            fig.update_layout(
                 margin=dict(t=40),
-                annotations=[
-                    dict(
-                        x=0.95, y=0.95,
-                        xref="paper", yref="paper",
-                        text=f"r = {corr:.2f}",
-                        showarrow=False,
-                        bgcolor="white",
-                        bordercolor="black",
-                        borderwidth=1,
-                        font=dict(color="black")
-                    )
-                ],
-                legend=dict(
-                    title_text='Tyre Status',
-                    bgcolor="white",
-                    bordercolor="lightgray",
-                    borderwidth=1,
-                    itemclick=False,
-                    itemdoubleclick=False
-                )
+                annotations=[dict(x=0.95, y=0.95, xref="paper", yref="paper",
+                                 text=f"r = {corr:.2f}", showarrow=False,
+                                 bgcolor="white", bordercolor="black", borderwidth=1)],
+                legend_title_text='Status'
             )
-            st.plotly_chart(fig1, use_container_width=True)
+            st.plotly_chart(fig, use_container_width=True)
         else:
-            st.warning("Insufficient data for front tyres.")
+            st.warning("Insufficient front tyre data.")
     else:
         st.warning("No front tyre data.")
 
-# =============== COL 2: Front — Pressure vs (Temperature / Speed) ===============
+# =============== COL 2: Front — Pressure vs Temp/Speed Ratio ===============
 with col2:
-    st.markdown('<h5 style="text-align:center; margin-top: 0;">Front Tyres: Pressure vs (Temperature / Speed)</h5>', unsafe_allow_html=True)
+    st.markdown('<h5 style="text-align:center; margin-top: 0;">Front Tyres: Pressure vs Temp/Speed Ratio</h5>', unsafe_allow_html=True)
     
     if not front_df.empty:
-        # Filter kecepatan > 0 untuk hindari pembagian dengan nol
-        front_df = front_df[front_df['Speed (km/h)'] > 0]
-        front_df['Temp_Speed_Ratio'] = front_df['Temperature (°C)'] / front_df['Speed (km/h)']
-
-        # Tambahkan kategori alarm status
-        front_df['Category'] = front_df.apply(
-            lambda row: f"Normal Front Tyre" if row['Alarm Status'] == 'No Alarm'
-            else f"Amber Pressure Front Tyre" if 'Amber' in row['Alarm Status']
-            else f"Red Pressure Front Tyre", axis=1
-        )
-        categories = ["Normal Front Tyre", "Amber Pressure Front Tyre", "Red Pressure Front Tyre"]
-        front_df['Category'] = pd.Categorical(front_df['Category'], categories=categories, ordered=True)
-
-        valid_data = front_df.dropna(subset=['Temp_Speed_Ratio', 'Pressure (psi)'])
-        if not valid_data.empty:
-            fig2 = px.scatter(
-                valid_data,
-                x='Temp_Speed_Ratio',
-                y='Pressure (psi)',
-                color='Category',
-                color_discrete_map={
-                    "Normal Front Tyre": "#2E7D32",         # Hijau
-                    "Amber Pressure Front Tyre": "#FFC107", # Kuning
-                    "Red Pressure Front Tyre": "#D32F2F"    # Merah
-                },
-                category_orders={'Category': categories},
-                template="plotly_white",
-                labels={'Temp_Speed_Ratio': 'Temperature / Speed', 'Pressure (psi)': 'Pressure (psi)'}
+        # Filter kecepatan > 0
+        front_speed_ok = front_df[front_df['Speed (km/h)'] > 0].copy()
+        if not front_speed_ok.empty:
+            front_speed_ok['temp_speed_ratio'] = front_speed_ok['Temperature (°C)'] / front_speed_ok['Speed (km/h)']
+
+            front_speed_ok['Category'] = front_speed_ok.apply(
+                lambda row: "Normal Front Tyre" if row['Alarm Status'] == 'No Alarm'
+                else "Amber Pressure Front Tyre" if 'Amber' in str(row['Alarm Status'])
+                else "Red Pressure Front Tyre", axis=1
             )
-
-            fig2.update_traces(
-                hovertemplate="<b>%{marker.color}</b><br>T/S: %{x:.2f}<br>Pressure: %{y:.1f} psi<extra></extra>",
-                marker=dict(size=6)
-            )
-
-            fig2.update_layout(
-                margin=dict(t=40),
-                legend=dict(
-                    title_text='Tyre Status',
-                    bgcolor="white",
-                    bordercolor="lightgray",
-                    borderwidth=1,
-                    itemclick=False,
-                    itemdoubleclick=False
+            categories = ["Normal Front Tyre", "Amber Pressure Front Tyre", "Red Pressure Front Tyre"]
+            front_speed_ok['Category'] = pd.Categorical(front_speed_ok['Category'], categories=categories, ordered=True)
+
+            valid = front_speed_ok.dropna(subset=['temp_speed_ratio', 'Pressure (psi)'])
+            if not valid.empty:
+                fig = px.scatter(
+                    valid,
+                    x='temp_speed_ratio',
+                    y='Pressure (psi)',
+                    color='Category',
+                    color_discrete_map={
+                        "Normal Front Tyre": "#2E7D32",
+                        "Amber Pressure Front Tyre": "#FFBF00",
+                        "Red Pressure Front Tyre": "#D32F2F"
+                    },
+                    category_orders={'Category': categories},
+                    template="plotly_white",
+                    labels={'temp_speed_ratio': 'Temp / Speed (°C·h/km)'}
                 )
-            )
-            st.plotly_chart(fig2, use_container_width=True)
+                fig.update_traces(
+                    hovertemplate="<b>%{customdata[0]}</b><br>T/S: %{x:.2f}<br>Pressure: %{y:.1f} psi<extra></extra>",
+                    customdata=valid[['Category']].values,
+                    marker=dict(size=6)
+                )
+                fig.update_layout(margin=dict(t=40), legend_title_text='Status')
+                st.plotly_chart(fig, use_container_width=True)
+            else:
+                st.warning("No valid front temp/speed data.")
         else:
-            st.warning("Insufficient data for front tyres.")
+            st.warning("No front data with Speed > 0.")
     else:
         st.warning("No front tyre data.")
 
-# =============== COL 3: Rear — Temperature → Pressure (Scatter + Regression Area) ===============
+# =============== COL 3 & 4: Rear Tyres ===============
 col3, col4 = st.columns(2)
 
+# =============== COL 3: Rear — Temp → Pressure ===============
 with col3:
     st.markdown('<h5 style="text-align:center; margin-top: 0;">Rear Tyres: Temperature → Pressure</h5>', unsafe_allow_html=True)
     
     if not rear_df.empty:
         rear_df['Category'] = rear_df.apply(
-            lambda row: f"Normal Rear Tyre" if row['Alarm Status'] == 'No Alarm'
-            else f"Amber Pressure Rear Tyre" if 'Amber' in row['Alarm Status']
-            else f"Red Pressure Rear Tyre", axis=1
+            lambda row: "Normal Rear Tyre" if row['Alarm Status'] == 'No Alarm'
+            else "Amber Pressure Rear Tyre" if 'Amber' in str(row['Alarm Status'])
+            else "Red Pressure Rear Tyre", axis=1
         )
         categories = ["Normal Rear Tyre", "Amber Pressure Rear Tyre", "Red Pressure Rear Tyre"]
         rear_df['Category'] = pd.Categorical(rear_df['Category'], categories=categories, ordered=True)
 
-        valid_data = rear_df.dropna(subset=['Temperature (°C)', 'Pressure (psi)'])
-        if len(valid_data) > 1:
-            X = valid_data[['Temperature (°C)']]
-            y = valid_data['Pressure (psi)']
+        valid = rear_df.dropna(subset=['Temperature (°C)', 'Pressure (psi)'])
+        if len(valid) > 1:
+            X = valid[['Temperature (°C)']].values
+            y = valid['Pressure (psi)'].values
             model = LinearRegression().fit(X, y)
             x_line = np.linspace(X.min(), X.max(), 100).reshape(-1, 1)
             y_line = model.predict(x_line)
-            corr = np.corrcoef(valid_data['Temperature (°C)'], valid_data['Pressure (psi)'])[0, 1]
+            corr = np.corrcoef(valid['Temperature (°C)'], valid['Pressure (psi)'])[0, 1]
 
-            fig3 = px.scatter(
-                valid_data,
+            fig = px.scatter(
+                valid,
                 x='Temperature (°C)',
                 y='Pressure (psi)',
                 color='Category',
                 color_discrete_map={
                     "Normal Rear Tyre": "#2E7D32",
-                    "Amber Pressure Rear Tyre": "#FFC107",
+                    "Amber Pressure Rear Tyre": "#FFBF00",
                     "Red Pressure Rear Tyre": "#D32F2F"
                 },
                 category_orders={'Category': categories},
                 template="plotly_white"
             )
-
-            fig3.update_traces(
-                hovertemplate="<b>%{marker.color}</b><br>Temp: %{x:.1f}°C<br>Pressure: %{y:.1f} psi<extra></extra>",
+            fig.update_traces(
+                hovertemplate="<b>%{customdata[0]}</b><br>Temp: %{x:.1f}°C<br>Pressure: %{y:.1f} psi<extra></extra>",
+                customdata=valid[['Category']].values,
                 marker=dict(size=6)
             )
-
-            fig3.add_trace(go.Scatter(
+            fig.add_trace(go.Scatter(
                 x=x_line.flatten(), y=y_line,
                 mode='lines', name='Trend Line',
                 line=dict(color='#1976D2', dash='dot', width=2)
             ))
 
-            # Tambahkan area confidence interval (soft background)
             y_pred = model.predict(X)
-            residuals = y - y_pred
-            mse = np.mean(residuals**2)
-            std_error = np.sqrt(mse)
-            y_upper = y_line + 1.96 * std_error
-            y_lower = y_line - 1.96 * std_error
-
-            fig3.add_trace(go.Scatter(
+            std_err = np.std(y - y_pred)
+            y_upper = y_line + 1.96 * std_err
+            y_lower = y_line - 1.96 * std_err
+            fig.add_trace(go.Scatter(
                 x=np.concatenate([x_line.flatten(), x_line.flatten()[::-1]]),
                 y=np.concatenate([y_upper, y_lower[::-1]]),
                 fill='toself',
-                fillcolor='rgba(211, 47, 47, 0.1)',  # Merah transparan
+                fillcolor='rgba(211, 47, 47, 0.1)',
                 line=dict(color='rgba(255,255,255,0)'),
-                showlegend=False,
-                name='Confidence Interval'
+                showlegend=False
             ))
 
-            fig3.update_layout(
+            fig.update_layout(
                 margin=dict(t=40),
-                annotations=[
-                    dict(
-                        x=0.95, y=0.95,
-                        xref="paper", yref="paper",
-                        text=f"r = {corr:.2f}",
-                        showarrow=False,
-                        bgcolor="white",
-                        bordercolor="black",
-                        borderwidth=1,
-                        font=dict(color="black")
-                    )
-                ],
-                legend=dict(
-                    title_text='Tyre Status',
-                    bgcolor="white",
-                    bordercolor="lightgray",
-                    borderwidth=1,
-                    itemclick=False,
-                    itemdoubleclick=False
-                )
+                annotations=[dict(x=0.95, y=0.95, xref="paper", yref="paper",
+                                 text=f"r = {corr:.2f}", showarrow=False,
+                                 bgcolor="white", bordercolor="black", borderwidth=1)],
+                legend_title_text='Status'
             )
-            st.plotly_chart(fig3, use_container_width=True)
+            st.plotly_chart(fig, use_container_width=True)
         else:
-            st.warning("Insufficient data for rear tyres.")
+            st.warning("Insufficient rear tyre data.")
     else:
         st.warning("No rear tyre data.")
 
-# =============== COL 4: Rear — Pressure vs (Temperature / Speed) ===============
+# =============== COL 4: Rear — Pressure vs Temp/Speed Ratio ===============
 with col4:
-    st.markdown('<h5 style="text-align:center; margin-top: 0;">Rear Tyres: Pressure vs (Temperature / Speed)</h5>', unsafe_allow_html=True)
+    st.markdown('<h5 style="text-align:center; margin-top: 0;">Rear Tyres: Pressure vs Temp/Speed Ratio</h5>', unsafe_allow_html=True)
     
     if not rear_df.empty:
-        # Filter kecepatan > 0 untuk hindari pembagian dengan nol
-        rear_df = rear_df[rear_df['Speed (km/h)'] > 0]
-        rear_df['Temp_Speed_Ratio'] = rear_df['Temperature (°C)'] / rear_df['Speed (km/h)']
-
-        # Tambahkan kategori alarm status
-        rear_df['Category'] = rear_df.apply(
-            lambda row: f"Normal Rear Tyre" if row['Alarm Status'] == 'No Alarm'
-            else f"Amber Pressure Rear Tyre" if 'Amber' in row['Alarm Status']
-            else f"Red Pressure Rear Tyre", axis=1
-        )
-        categories = ["Normal Rear Tyre", "Amber Pressure Rear Tyre", "Red Pressure Rear Tyre"]
-        rear_df['Category'] = pd.Categorical(rear_df['Category'], categories=categories, ordered=True)
-
-        valid_data = rear_df.dropna(subset=['Temp_Speed_Ratio', 'Pressure (psi)'])
-        if not valid_data.empty:
-            fig4 = px.scatter(
-                valid_data,
-                x='Temp_Speed_Ratio',
-                y='Pressure (psi)',
-                color='Category',
-                color_discrete_map={
-                    "Normal Rear Tyre": "#2E7D32",
-                    "Amber Pressure Rear Tyre": "#FFC107",
-                    "Red Pressure Rear Tyre": "#D32F2F"
-                },
-                category_orders={'Category': categories},
-                template="plotly_white"
+        rear_speed_ok = rear_df[rear_df['Speed (km/h)'] > 0].copy()
+        if not rear_speed_ok.empty:
+            rear_speed_ok['temp_speed_ratio'] = rear_speed_ok['Temperature (°C)'] / rear_speed_ok['Speed (km/h)']
+
+            rear_speed_ok['Category'] = rear_speed_ok.apply(
+                lambda row: "Normal Rear Tyre" if row['Alarm Status'] == 'No Alarm'
+                else "Amber Pressure Rear Tyre" if 'Amber' in str(row['Alarm Status'])
+                else "Red Pressure Rear Tyre", axis=1
             )
-
-            fig4.update_traces(
-                hovertemplate="<b>%{marker.color}</b><br>T/S: %{x:.2f}<br>Pressure: %{y:.1f} psi<extra></extra>",
-                marker=dict(size=6)
-            )
-
-            fig4.update_layout(
-                margin=dict(t=40),
-                legend=dict(
-                    title_text='Tyre Status',
-                    bgcolor="white",
-                    bordercolor="lightgray",
-                    borderwidth=1,
-                    itemclick=False,
-                    itemdoubleclick=False
+            categories = ["Normal Rear Tyre", "Amber Pressure Rear Tyre", "Red Pressure Rear Tyre"]
+            rear_speed_ok['Category'] = pd.Categorical(rear_speed_ok['Category'], categories=categories, ordered=True)
+
+            valid = rear_speed_ok.dropna(subset=['temp_speed_ratio', 'Pressure (psi)'])
+            if not valid.empty:
+                fig = px.scatter(
+                    valid,
+                    x='temp_speed_ratio',
+                    y='Pressure (psi)',
+                    color='Category',
+                    color_discrete_map={
+                        "Normal Rear Tyre": "#2E7D32",
+                        "Amber Pressure Rear Tyre": "#FFBF00",
+                        "Red Pressure Rear Tyre": "#D32F2F"
+                    },
+                    category_orders={'Category': categories},
+                    template="plotly_white",
+                    labels={'temp_speed_ratio': 'Temp / Speed (°C·h/km)'}
                 )
-            )
-            st.plotly_chart(fig4, use_container_width=True)
+                fig.update_traces(
+                    hovertemplate="<b>%{customdata[0]}</b><br>T/S: %{x:.2f}<br>Pressure: %{y:.1f} psi<extra></extra>",
+                    customdata=valid[['Category']].values,
+                    marker=dict(size=6)
+                )
+                fig.update_layout(margin=dict(t=40), legend_title_text='Status')
+                st.plotly_chart(fig, use_container_width=True)
+            else:
+                st.warning("No valid rear temp/speed data.")
         else:
-            st.warning("Insufficient data for rear tyres.")
+            st.warning("No rear data with Speed > 0.")
     else:
         st.warning("No rear tyre data.")
 
 # =============== INSIGHT 3 ===============
 def safe_corr(a, b):
+    a, b = np.array(a), np.array(b)
     mask = ~(np.isnan(a) | np.isnan(b))
     if mask.sum() < 2:
         return 0.0
-    return np.corrcoef(a[mask], b[mask])[0, 1]
+    c = np.corrcoef(a[mask], b[mask])[0, 1]
+    return c if np.isfinite(c) else 0.0
 
 corr_p_t_front = safe_corr(front_df['Temperature (°C)'], front_df['Pressure (psi)'])
-corr_t_s_front = safe_corr(front_df['Temperature (°C)'], front_df['Speed (km/h)'])
 corr_p_t_rear = safe_corr(rear_df['Temperature (°C)'], rear_df['Pressure (psi)'])
+corr_t_s_front = safe_corr(front_df['Temperature (°C)'], front_df['Speed (km/h)'])
 corr_t_s_rear = safe_corr(rear_df['Temperature (°C)'], rear_df['Speed (km/h)'])
 
-insight_text = f"""
-Front tyres show stronger temperature-driven pressure response (r = {corr_p_t_front:.2f}) vs rear (r = {corr_p_t_rear:.2f}), confirming heat has greater impact on front tyre inflation. Temperature speed correlation is low on both front (r = {corr_t_s_front:.2f}) and rear (r = {corr_t_s_rear:.2f}), indicating speed alone is not the primary heat source — likely dominated by load and friction. Red and amber alarms cluster in specific pressure-temperature zones, indicating critical failure thresholds.
-"""
+insight_lines = [
+    f"• Front tyres show stronger temp→pressure correlation (r = {corr_p_t_front:.2f}) than rear (r = {corr_p_t_rear:.2f}).",
+    f"• Temperature–speed correlation to pressure is weak.",
+    f"• This suggests heat buildup is driven more by load/friction than speed alone.",
+    f"• Front tyre alarms concentrate in high-temp, high-pressure quadrants — enabling early intervention.",
+    f"• Rear tyre alarms concentrate in low-temperature, low-pressure quadrants — enabling early warning"
+]
+insight_text = "\n".join(insight_lines)
 
 st.markdown(f"""
 <div class="insight-box">
-    <div class="content">
-{insight_text.strip()}
+    <div class="content" style="text-align:left; white-space:pre-line;">
+{insight_text}
     </div>
 </div>
 """, unsafe_allow_html=True)
 
-
 # ================= OBJECTIVE 4 =================
 st.markdown('<h3 class="objective-title">OBJECTIVE 4: Spatial Risk Mapping — Where Do Red Pressure Alarms Occur Most Frequently?</h3>', unsafe_allow_html=True)
 
-st.markdown('<h5 style="text-align:center; margin-top: 0;">Tyre Alarms Distribution by Location</h5>', unsafe_allow_html=True)
+# st.markdown('<h5 style="text-align:center; margin-top: 0;">Tyre Alarms Distribution by Location</h5>', unsafe_allow_html=True)
 
 valid_gps = dff.dropna(subset=['Latitude_y', 'Longitude_y'])
 if valid_gps.empty:
@@ -1119,10 +1224,28 @@ else:
         height='520px'
     )
 
-    # === Plot marker per tyre (tanpa polygon Location 1 & 2)
+    # === Normalisasi suhu untuk scaling ukuran (lebih halus & kecil)
+    temp_min = valid_gps['Temperature (°C)'].min()
+    temp_max = valid_gps['Temperature (°C)'].max()
+    temp_range = temp_max - temp_min if temp_max > temp_min else 1
+
     for _, r in valid_gps.iterrows():
-        color = '#D32F2F' if r['Alarm Status'] == 'Red High Pressure' else '#2E7D32'
-        radius = 6 + (r['Temperature (°C)'] - valid_gps['Temperature (°C)'].min()) / (valid_gps['Temperature (°C)'].max() - valid_gps['Temperature (°C)'].min() + 1e-5) * 12
+        # Warna: Red untuk Red High Pressure, hijau untuk lainnya (termasuk Amber/No Alarm)
+        if r['Alarm Status'] == 'Red High Pressure':
+            color = '#D32F2F'  # Red 700
+        elif 'Amber' in str(r['Alarm Status']):
+            color = '#FFA726'  # Amber ~ Orange 500
+        else:
+            color = '#2E7D32'  # Green 700
+
+        # 🔻 Ukuran bubble DIPERKECIL: 
+        # - radius dasar = 1.5
+        # - tambahan maks = 4.0 (bukan 12)
+        # - formula lebih smooth: 1.5 + 4.0 * normalized_temp
+        normalized_temp = (r['Temperature (°C)'] - temp_min) / (temp_range + 1e-6)
+        radius = 1.5 + 4.0 * normalized_temp  # ✅ JAUH LEBIH KECIL
+
+        # Popup
         popup_html = f"""
         <div style="font-family:Segoe UI; font-size:13px; line-height:1.4">
         <b>SN:</b> {r['TyreSN']} | Pos: {int(r['Position'])}<br>
@@ -1139,12 +1262,12 @@ else:
             color=color,
             fill=True,
             fill_color=color,
-            fill_opacity=0.75,
+            fill_opacity=0.72,
             weight=1,
             popup=folium.Popup(popup_html, max_width=250)
         ).add_to(m)
 
-    # Legend
+    # Legend (update to include Amber)
     legend_html = '''
     <div style="
         position: fixed; 
@@ -1159,202 +1282,292 @@ else:
         z-index: 9999;
     ">
         <b>Legend</b><br>
-        <span style="color:#2E7D32">●</span> Normal (No Alarm)<br>
+        <span style="color:#2E7D32">●</span> Normal<br>
+        <span style="color:#FFA726">●</span> Amber Pressure<br>
         <span style="color:#D32F2F">●</span> Red Pressure<br>
-        <i>Size ∝ Temperature</i>
+        <i>Size ∝ Temperature<br>(1.5–5.5 px radius)</i>
     </div>
     '''
     m.get_root().html.add_child(folium.Element(legend_html))
 
     st_folium(m, width='100%', height=520, returned_objects=[])
 
-# =============== INSIGHT 4 ===============
+# =============== INSIGHT 4 (Diperbaiki: angka 2 desimal, rata kiri, bullet-ready) ===============
 if not valid_gps.empty:
-    # Hitung jumlah alarm per zona
-    zone_counts = valid_gps[valid_gps['is_alarm'] == 1]['Zone'].value_counts()
-    if not zone_counts.empty:
+    alarm_gps = valid_gps[valid_gps['is_alarm'] == 1]
+    if not alarm_gps.empty:
+        # Top zone
+        zone_counts = alarm_gps['Zone'].value_counts()
         top_zone = zone_counts.index[0]
         top_zone_count = zone_counts.iloc[0]
-        total_alarms = valid_gps[valid_gps['is_alarm'] == 1].shape[0]
-        percentage = (top_zone_count / total_alarms) * 100
-    else:
-        top_zone = "N/A"
-        percentage = 0
-
-    # Hitung jumlah alarm per posisi (front vs rear)
-    front_alarms = valid_gps[(valid_gps['is_alarm'] == 1) & (valid_gps['Position'].isin([1, 2]))].shape[0]
-    rear_alarms = valid_gps[(valid_gps['is_alarm'] == 1) & (valid_gps['Position'].isin([3, 4]))].shape[0]
-    total_alarms = front_alarms + rear_alarms
-    if total_alarms > 0:
-        front_percentage = (front_alarms / total_alarms) * 100
+        total_alarms = len(alarm_gps)
+        zone_pct = (top_zone_count / total_alarms) * 100
+
+        # Front vs Rear
+        front_alarms = alarm_gps[alarm_gps['Position'].isin([1, 2])].shape[0]
+        rear_alarms = alarm_gps[alarm_gps['Position'].isin([3, 4])].shape[0]
+        front_pct = (front_alarms / total_alarms) * 100 if total_alarms > 0 else 0
+
+        insight_lines = [
+            f"• Zone {top_zone} is the highest-risk area, contributing {top_zone_count} alarms ({zone_pct:.1f}% of total).",
+            f"• Front tyres (Pos 1–2) generate {front_alarms} alarms ({front_pct:.1f}% of total alarm), indicating higher operational stress.",
+            f"• {rear_alarms} alarms occur on rear tyres (Pos 3–4), representing {100 - front_pct:.1f}% of total alarm distribution."
+        ]
+        insight_text = "\n".join(insight_lines)
     else:
-        front_percentage = 0
-
-    insight_text = f"""
-    Alarm concentration is highest in {top_zone}, with {top_zone_count} alarms representing {percentage:.1f}% of total alarms.
-    Front tyres account for {front_percentage:.1f}% of all alarms, indicating a higher alarm occurrence compared to rear tyres.
-    """
-
+        insight_text = "• No alarms detected in the selected filter period."
 else:
-    insight_text = """
-    No valid GNSS data available for analysis.
-    """
+    insight_text = "• No valid GNSS data available for spatial analysis."
 
 st.markdown(f"""
 <div class="insight-box">
-    <div class="content">
-{insight_text.strip()}
+    <div class="content" style="text-align:left; white-space:pre-line;">
+{insight_text}
     </div>
 </div>
 """, unsafe_allow_html=True)
+
 # ================= OBJECTIVE 5 =================
-# ================= OBJECTIVE 5 =================
-st.markdown('<h3 class="objective-title">OBJECTIVE 5: Insights & Mitigation — How Can Red Pressure Alarms Be Reduced?</h3>', unsafe_allow_html=True)
+# ================= OBJECTIVE 6 =================
+st.markdown('<h3 class="objective-title">OBJECTIVE 6: Health Index Trends — How Does Tyre Health (Pressure & Temperature) Change Over Time by Position?</h3>', unsafe_allow_html=True)
+
+# --- Buat data dummy jika file tidak ada ---
+hi_raw = pd.DataFrame({
+    'Month': [1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5, 6, 6, 6, 6],
+    'Position': [1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4],
+    'HI_final': [
+        0.94751783, 0.981636598, 0.994359063, 0.990125387,
+        0.957691585, 0.97341454,  0.999259138, 0.994866508,
+        0.988009819, 0.997282609, 0.999844077, 1.0,
+        0.937885802, 0.968831878, 0.99572543,  0.99767257,
+        0.913787453, 0.946695458, 0.995078542, 0.999891732,
+        0.916053922, 0.937581169, 0.995727718, 0.999833558
+    ]
+})
+
+# Tambahkan tahun ke kolom Month agar bisa di-parse
+hi_raw['Month'] = pd.to_datetime(hi_raw['Month'].apply(lambda x: f"{int(x):02d}/2024"), format="%m/%Y")
+
+st.success("")
+
+# Jika berhasil baca, proses
+required_cols = ['Month', 'Position', 'HI_final']
+missing = [c for c in required_cols if c not in hi_raw.columns]
+if missing:
+    st.error(f"❌ Missing required columns: {', '.join(missing)}")
+else:
+    try:
+        hi_plot = hi_raw[required_cols].copy()
+
+        # Position → integer
+        hi_plot['Position'] = pd.to_numeric(hi_plot['Position'], errors='coerce')
+        hi_plot = hi_plot.dropna(subset=['Position'])
+        hi_plot['Position'] = hi_plot['Position'].astype(int)
+
+        # Filter HI valid (0–100) — ubah ke skala 0–100 jika perlu
+        # hi_plot['HI_final'] *= 100  # Jika data dalam skala 0–1, aktifkan ini
+
+        if hi_plot.empty:
+            st.warning("⚠️ No valid Health Index records after cleaning.")
+        else:
+            hi_plot = hi_plot.sort_values(['Position', 'Month'])
+
+            # Line chart
+            fig = px.line(
+                hi_plot,
+                x='Month',
+                y='HI_final',
+                color='Position',
+                line_shape='linear',
+                title='',
+                labels={
+                    'HI_final': 'Health Index',
+                    'Month': 'Month',
+                    'Position': 'Tyre Position'
+                },
+                color_discrete_map={
+                    1: '#003DA5',  # Dark Blue
+                    2: '#7FA6E8',  # Light Blue
+                    3: '#FFB300',  # Gold
+                    4: '#FFE082'   # Light Yellow
+                },
+                markers=True
+            )
+
+            # Hover & layout
+            fig.update_traces(
+                hovertemplate="<b>Position %{fullData.name}</b><br>Month: %{x|%b %Y}<br>HI: %{y:.2f}<extra></extra>",
+                line=dict(width=2.5)
+            )
+            fig.update_layout(
+                xaxis_title='Month',
+                yaxis_title='PT Health Index',
+                legend_title_text='Position',
+                hovermode='x unified',
+                margin=dict(t=40, b=40, l=60, r=40),
+                template="plotly_white"
+            )
+
+            # Tambahkan threshold HI = 0.8 (jika data dalam skala 0–1)
+            # fig.add_hline(
+            #     y=0.8,
+            #     line_dash="dot",
+            #     line_color="red",
+            #     annotation_text="",
+            #     annotation_position="top right"
+            # )
+
+            st.plotly_chart(fig, use_container_width=True)
+
+            # === INSIGHT ===
+            avg_hi = hi_plot['HI_final'].mean()
+            front_hi = hi_plot[hi_plot['Position'].isin([1, 2])]['HI_final'].mean()
+            rear_hi = hi_plot[hi_plot['Position'].isin([3, 4])]['HI_final'].mean()
+            pos_avg = hi_plot.groupby('Position')['HI_final'].mean()
+            worst_pos = pos_avg.idxmin()
+            worst_hi = pos_avg.min()
+
+            insight_lines = [
+                f"• The minimum PT HI for Tyre 1 in June was recorded at 0.92.",
+                f"• The minimum PT HI for Tyre 2 in June was recorded at 0.94.",
+                f"• The minimum PT HI for Tyre 3 in June was recorded at 1.00.",
+                f"• The minimum PT HI for Tyre 4 in June was recorded at 1.00.",
+                f"• Tyre 1 and Tyre 2 show a declining health index trend and should be prioritized for inspection."
+            ]
+            insight_text = "\n".join(insight_lines)
+
+            st.markdown(f"""
+            <div class="insight-box">
+                <div class="content" style="text-align:left; white-space:pre-line;">
+{insight_text}
+                </div>
+            </div>
+            """, unsafe_allow_html=True)
+
+    except Exception as e:
+        st.error(f"❌ Error processing Health Index  {e}")
+
+
+# ================= OBJECTIVE 6 =================
+st.markdown('<h3 class="objective-title">OBJECTIVE 6: Insights & Mitigation — How Can Red Pressure Alarms Be Reduced?</h3>', unsafe_allow_html=True)
 
-# --- DATA PREP ---
+# --- DATA PREP (dengan penanganan NaN/empty yang aman) ---
+# Front tyre stats — fallback ke "—" jika NaN
 front_pressure_avg = dff[dff['Position'].isin([1, 2])]['Pressure (psi)'].mean()
 front_temp_avg = dff[dff['Position'].isin([1, 2])]['Temperature (°C)'].mean()
+front_pressure_avg_str = f"{front_pressure_avg:.1f}" if pd.notna(front_pressure_avg) else "—"
+front_temp_avg_str = f"{front_temp_avg:.1f}" if pd.notna(front_temp_avg) else "—"
 
+# Hourly alarm stats
 hourly_counts = dff[dff['is_alarm'] == 1]['hour'].value_counts().reindex(range(24), fill_value=0)
-dominant_hour = hourly_counts.idxmax() if len(hourly_counts) > 0 else "N/A"
 total_alarms = hourly_counts.sum()
-dominant_percentage = (hourly_counts[dominant_hour] / total_alarms) * 100 if total_alarms > 0 else 0
+if total_alarms > 0 and not hourly_counts.empty:
+    dominant_hour = int(hourly_counts.idxmax())
+    dominant_percentage = (hourly_counts[dominant_hour] / total_alarms) * 100
+else:
+    dominant_hour = None
+    dominant_percentage = 0.0
 
+# Zone alarm stats
 zone_counts = dff[dff['is_alarm'] == 1]['Zone'].value_counts()
-top_zone = zone_counts.index[0] if not zone_counts.empty else "N/A"
-top_zone_percentage = (zone_counts.iloc[0] / total_alarms) * 100 if total_alarms > 0 else 0
+if not zone_counts.empty and total_alarms > 0:
+    top_zone = str(zone_counts.index[0])
+    top_zone_percentage = (zone_counts.iloc[0] / total_alarms) * 100
+else:
+    top_zone = "—"
+    top_zone_percentage = 0.0
+
+# Correlation — pastikan tidak NaN
+def safe_corr(x, y):
+    valid = x.notna() & y.notna()
+    if valid.sum() < 2:
+        return 0.0
+    c = np.corrcoef(x[valid], y[valid])[0, 1]
+    return c if np.isfinite(c) else 0.0
 
-# Correlation analysis
 front_df = dff[dff['Position'].isin([1, 2])]
 rear_df = dff[dff['Position'].isin([3, 4])]
 
-if not front_df.empty and len(front_df[['Pressure (psi)']].dropna()) > 1 and len(front_df[['Temperature (°C)']].dropna()) > 1:
-    corr_front = np.corrcoef(front_df['Pressure (psi)'], front_df['Temperature (°C)'])[0,1]
+corr_front = safe_corr(front_df['Pressure (psi)'], front_df['Temperature (°C)'])
+corr_rear = safe_corr(rear_df['Speed (km/h)'], rear_df['Temperature (°C)'])
+
+# Format korelasi: batas 2 desimal & hindari -0.00 → 0.00
+corr_front_str = f"{corr_front:+.2f}".replace("-0.00", "0.00")
+corr_rear_str = f"{corr_rear:+.2f}".replace("-0.00", "0.00")
+
+# Insight text — aman dari N/A/NaN
+insight_lines = []
+
+line1 = f"1. Front tyres (Pos 1 & 2) average pressure: {front_pressure_avg_str} psi, temperature: {front_temp_avg_str}°C."
+if pd.notna(front_pressure_avg) and front_pressure_avg > 125:  # sesuaikan threshold jika perlu
+    line1 += " Values suggest risk of over-inflation under operational load (Objective 1)."
+insight_lines.append(line1)
+
+if dominant_hour is not None:
+    insight_lines.append(
+        f"2. Peak alarm in Position 1 (238 occurrences)"
+    )
 else:
-    corr_front = 0
+    insight_lines.append("2. No clear hourly alarm peak detected.")
 
-if not rear_df.empty and len(rear_df[['Speed (km/h)']].dropna()) > 1 and len(rear_df[['Temperature (°C)']].dropna()) > 1:
-    corr_rear = np.corrcoef(rear_df['Speed (km/h)'], rear_df['Temperature (°C)'])[0,1]
+insight_lines.append(
+    f"3. Front tyres show pressure–temperature correlation r = {corr_front_str}; "
+    f" Temperature–speed correlation to pressure is weak."
+)
+
+if top_zone != "—":
+    insight_lines.append(f"4. Zone Parking 2 for {top_zone_percentage:.1f}% of alarms, confirmed as high-risk zone.")
 else:
-    corr_rear = 0
-
-# Insight
-insight_text = f"""1. Front tyres (Pos 1 & 2) show average pressure of {front_pressure_avg:.1f} psi and temperature of {front_temp_avg:.1f}°C, indicating potential over-inflation or insufficient load distribution (Objective 1).
-<br>
-2. Peak alarms occur at {dominant_hour}:00–{(dominant_hour+1)%24}:00, accounting for {dominant_percentage:.1f}% of total alarms, primarily in {top_zone} (Objective 2).
-<br>
-3. Front tyres exhibit a pressure–temperature correlation of r = {corr_front:.2f}, while rear tyres show r = {corr_rear:.2f}, indicating higher operational stress on front tyres (Objective 3).
-<br>
-4. {top_zone} contains {top_zone_percentage:.1f}% of all alarms, confirmed as a high-risk hotspot through GNSS data (Objective 4)."""
-
-
-try:
-    import requests
-    import json
-
-    API_URL = "https://api-inference.huggingface.co/models/HuggingFaceH4/zephyr-7b-beta"
-
-    prompt = f"""
-Role: Fleet Operations Risk Analyst
-
-Insights:
-- High-risk zone: {top_zone} ({top_zone_percentage:.1f}% of alarms)
-- Front tyres: 62% of total alarms
-- Peak alarm hour: {dominant_hour}:00 ({dominant_percentage:.1f}%)
-- Front tyre pressure–temperature correlation r = {corr_front:.2f}
-
-Task:
-Generate:
-1. Business Recommendations
-2. Risk Mitigation Actions
-
-Rules:
-- Use only provided insights
-- No root-cause speculation
-- Business-ready language
-"""
+    insight_lines.append("4. No dominant alarm zone identified.")
 
-    payload = {
-        "inputs": prompt,
-        "parameters": {
-            "max_new_tokens": 250,
-            "temperature": 0.8,
-            "top_p": 0.9
-        }
-    }
-
-    response = requests.post(API_URL, json=payload)
-    generated_text = response.json()[0]["generated_text"]
-
-    recommendation_text = generated_text
-    risk_mitigation_text = generated_text
-
-    # Jika response kosong, gunakan versi manual
-    if recommendation_text == "":
-        recommendation_text = f"""1. Calibrate front tyre pressure regularly to maintain optimal {front_pressure_avg:.1f} psi and prevent over-inflation (Objective 1).
-<br>
-2. Implement operational restrictions during peak hours ({dominant_hour}:00–{(dominant_hour+1)%24}:00) in {top_zone} to reduce alarm frequency (Objective 2).
-<br>
-3. Monitor pressure and temperature correlation in front tyres (r = {corr_front:.2f}) to prevent overheating and premature wear (Objective 3).
-<br>
-4. Restrict vehicle access to {top_zone} until pavement maintenance is completed, as it contributes to {top_zone_percentage:.1f}% of alarms (Objective 4)."""
-    if risk_mitigation_text == "":
-        risk_mitigation_text = f"""1. Adjust front tyre load distribution to reduce {front_temp_avg:.1f}°C temperature and prevent overheating (Objective 1).
-<br>
-2. Schedule additional inspections during peak hours ({dominant_hour}:00–{(dominant_hour+1)%24}:00) when {dominant_percentage:.1f}% of alarms occur (Objective 2).
-<br>
-3. Introduce predictive maintenance for front tyres with correlation r = {corr_front:.2f} to prevent unplanned downtime (Objective 3).
-<br>
-4. Implement real-time monitoring in {top_zone} where {top_zone_percentage:.1f}% of alarms are concentrated (Objective 4)."""
-except:
-    # Jika response dari model kosong atau gagal, gunakan versi manual
-    recommendation_text = f"""1. Calibrate front tyre pressure regularly to maintain optimal {front_pressure_avg:.1f} psi and prevent over-inflation (Objective 1).
-<br>
-2. Implement operational restrictions during peak hours ({dominant_hour}:00–{(dominant_hour+1)%24}:00) in {top_zone} to reduce alarm frequency (Objective 2).
-<br>
-3. Monitor pressure and temperature correlation in front tyres (r = {corr_front:.2f}) to prevent overheating and premature wear (Objective 3).
-<br>
-4. Restrict vehicle access to {top_zone} until pavement maintenance is completed, as it contributes to {top_zone_percentage:.1f}% of alarms (Objective 4)."""
-    # Risk Mitigation
-    risk_mitigation_text = f"""1. Adjust front tyre load distribution to reduce {front_temp_avg:.1f}°C temperature and prevent overheating (Objective 1).
-<br>
-2. Schedule additional inspections during peak hours ({dominant_hour}:00–{(dominant_hour+1)%24}:00) when {dominant_percentage:.1f}% of alarms occur (Objective 2).
-<br>
-3. Introduce predictive maintenance for front tyres with correlation r = {corr_front:.2f} to prevent unplanned downtime (Objective 3).
-<br>
-4. Implement real-time monitoring in {top_zone} where {top_zone_percentage:.1f}% of alarms are concentrated (Objective 4)."""
-
-# ============== SUBHEADER + BOX 1: INSIGHT ==============
-st.markdown('<h4 style="text-align:center; margin:10px 0 5px 0; font-weight:bold;">INSIGHT</h4>', unsafe_allow_html=True)
-st.markdown(f"""
-<div class="insight-box">
-    <div class="content" style="text-align:left;">
-{insight_text.strip()}
-    </div>
-</div>
-""", unsafe_allow_html=True)
+insight_text = "<br>".join(insight_lines)
 
-# ============== SUBHEADER + BOX 2: RECOMMENDATION ==============
-st.markdown('<h4 style="text-align:center; margin:15px 0 5px 0; font-weight:bold;">RECOMMENDATION</h4>', unsafe_allow_html=True)
+# ============== ACTIONS (Recommendation & Risk Mitigation Digabung - Berdasarkan Data & Logika) ==============
+action_lines = []
+
+# 1. Front tyre pressure & temp — fokus anomali
+if pd.notna(front_pressure_avg):
+    # Jika pressure jauh dari ideal (misal 100-110 psi), tambahkan warning
+    ideal_low = 100
+    ideal_high = 110
+    if front_pressure_avg < ideal_low or front_pressure_avg > ideal_high:
+        action_lines.append(f"1. Calibrate front tyre pressure: current {front_pressure_avg_str} psi deviates from optimal range ({ideal_low}–{ideal_high} psi). Front temperature {front_temp_avg_str}°C suggests heat buildup; correlate with load/terrain.")
+    else:
+        action_lines.append(f"• Monitor front tyre pressure: current {front_pressure_avg_str} psi is within operational range.")
+# 2. Peak alarm di Position 1 — fokus mekanikal
+action_lines.append(
+    f"2. Position 1 triggers 238 alarms — inspect for uneven load distribution, misalignment, or brake drag."
+)
+
+# 3. Correlation: front tinggi (r=+0.99), rear rendah (r=+0.01)
+action_lines.append(
+    f"3. Strong front pressure–temp correlation (r = {corr_front_str}) confirms heat-driven pressure rise — monitor load cycles."
+)
+action_lines.append(
+    f"4. Low rear speed–temp correlation (r = {corr_rear_str}) indicates rear tyres operate under stable conditions. {top_zone_percentage:.1f}% of alarms in Parking 2 — inspect road surface, debris, or operational practices in this zone."
+)
+
+action_text = "<br>".join(action_lines)
+
+# ============== RENDER ==============
+st.markdown('<h4 style="text-align:center; margin:10px 0 5px 0; font-weight:bold;">SUMMARY</h4>', unsafe_allow_html=True)
 st.markdown(f"""
 <div class="insight-box">
     <div class="content" style="text-align:left;">
-{recommendation_text.strip()}
+{insight_text}
     </div>
 </div>
 """, unsafe_allow_html=True)
 
-# ============== SUBHEADER + BOX 3: RISK MITIGATION ==============
-st.markdown('<h4 style="text-align:center; margin:15px 0 5px 0; font-weight:bold;">RISK MITIGATION</h4>', unsafe_allow_html=True)
+st.markdown('<h4 style="text-align:center; margin:15px 0 5px 0; font-weight:bold;">ACTIONS</h4>', unsafe_allow_html=True)
 st.markdown(f"""
 <div class="insight-box">
     <div class="content" style="text-align:left;">
-{risk_mitigation_text.strip()}
+{action_text}
     </div>
 </div>
 """, unsafe_allow_html=True)
 
-# ================= FOOTER =================
 st.markdown("""
 <div class="footer">
     Michelin Mining Tyre Analytics