Update app.py

app.py

@@ -1156,7 +1156,7 @@ import streamlit as st
 import pandas as pd
 import numpy as np
 import plotly.graph_objects as go
-st.subheader("OBJECTIVE 5: See your team’s fatigue Fatigue Hazard Profile!")
+st.subheader("OBJECTIVE 5: See your team’s Fatigue Hazard Profile!")
 
 # Custom CSS — tetap seperti sebelumnya (sudah sesuai preferensi)
 st.markdown("""
@@ -1384,7 +1384,7 @@ else:
         # ===============================================================
         # LEGEND — UPDATED: Stable → One Time Event, Gray → Yellow
         # ===============================================================
-        st.subheader("Hazard Gradient Legend")
+        st.subheader("Legend of Frequency Trends")
 
         st.markdown("""
         <div class="legend-container">
@@ -1519,15 +1519,23 @@ else:
             )
             return fig
 
+        # ===============================================================
+        # CHARTS
+        # ===============================================================
         # ===============================================================
         # CHARTS
         # ===============================================================
         col1, col2 = st.columns(2)
         with col1:
-            st.plotly_chart(plot_chart(top_ob, "OB HAULER Operators (Hazard Gradient)"), use_container_width=True)
+            # Center-align chart
+            st.markdown("<div style='display: flex; justify-content: center;'>", unsafe_allow_html=True)
+            st.plotly_chart(plot_chart(top_ob, "OB HAULER Operator Hazard Profile"), use_container_width=False, config={'displayModeBar': False})
+            st.markdown("</div>", unsafe_allow_html=True)
         with col2:
-            st.plotly_chart(plot_chart(top_coal, "HAULING COAL Operators (Hazard Gradient)"), use_container_width=True)
-
+            # Center-align chart
+            st.markdown("<div style='display: flex; justify-content: center;'>", unsafe_allow_html=True)
+            st.plotly_chart(plot_chart(top_coal, "COAL HAULING Operator Hazard Profile"), use_container_width=False, config={'displayModeBar': False})
+            st.markdown("</div>", unsafe_allow_html=True)
         # ===============================================================
         # AI INSIGHTS — DIPERBAIKI: Risk Summary jadi 1 box + 3 list
         # ===============================================================
@@ -1752,7 +1760,6 @@ with col_insights:
 
     # ===================== 2. High-Speed Fatigue Analysis =====================
     if col_speed and col_speed in df.columns:
-
         high_speed_threshold = 20
         high_speed_fatigue = df[df[col_speed] >= high_speed_threshold]
         high_speed_pct = (len(high_speed_fatigue) / len(df)) * 100 if len(df) > 0 else 0
@@ -1775,10 +1782,221 @@ with col_insights:
             st.info(
                 f"{high_speed_pct:.1f}% of alerts occur at high speeds. This is within acceptable range."
             )
-
     else:
         st.info("Speed data not available for High-Speed Fatigue Analysis.")
 
+    # ===================== 3. Weekend vs Weekday Risk =====================
+    df['is_weekend'] = df['start'].dt.weekday >= 5
+    weekend_alerts = len(df[df['is_weekend']])
+    weekday_alerts = len(df[~df['is_weekend']])
+    weekend_pct = (weekend_alerts / len(df)) * 100 if len(df) > 0 else 0
+
+    st.markdown(f"**Weekend vs Weekday Risk Distribution**")
+    st.markdown(
+        f"""
+        <div style="display: flex; gap: 10px; justify-content: space-between;">
+            <div style="text-align: center; flex: 1;">
+                <div style="font-size: 18px; font-weight: bold;">{weekend_alerts}</div>
+                <div style="color: #555;">Weekend</div>
+                <div style="color: red; font-size: 12px;">{weekend_pct:.1f}%</div>
+            </div>
+            <div style="text-align: center; flex: 1;">
+                <div style="font-size: 18px; font-weight: bold;">{weekday_alerts}</div>
+                <div style="color: #555;">Weekday</div>
+                <div style="color: green; font-size: 12px;">{100-weekend_pct:.1f}%</div>
+            </div>
+        </div>
+        """,
+        unsafe_allow_html=True
+    )
+    if weekend_pct > 60:
+        st.warning(f"High weekend fatigue risk ({weekend_pct:.1f}%). Consider weekend rest protocols.")
+    else:
+        st.info(f"Weekend fatigue is {weekend_pct:.1f}% — balanced with weekday activity.")
+
+    # ===================== 4. Shift-Based Risk Analysis =====================
+    if col_shift and col_shift in df.columns:
+        shift_counts = df[col_shift].value_counts()
+        st.markdown(f"**Shift-Based Risk Distribution**")
+        for shift, count in shift_counts.items():
+            shift_pct = (count / len(df)) * 100
+            st.markdown(f"- **Shift {shift}**: {count} alerts ({shift_pct:.1f}%)")
+        max_shift = shift_counts.idxmax() if not shift_counts.empty else "N/A"
+        if max_shift != "N/A":
+            st.warning(f"Shift {max_shift} has the highest fatigue alert count. Investigate scheduling or rest breaks.")
+    else:
+        st.info("Shift data not available for analysis.")
+
+    # ===================== 5. Fleet Type Risk Distribution =====================
+    if col_fleet_type and col_fleet_type in df.columns:
+        fleet_counts = df[col_fleet_type].value_counts()
+        st.markdown(f"**Fleet Type Risk Distribution**")
+        for fleet, count in fleet_counts.items():
+            fleet_pct = (count / len(df)) * 100
+            st.markdown(f"- **{fleet}**: {count} alerts ({fleet_pct:.1f}%)")
+        max_fleet = fleet_counts.idxmax() if not fleet_counts.empty else "N/A"
+        if max_fleet != "N/A":
+            st.warning(f"Fleet type '{max_fleet}' has the highest fatigue alert count. Investigate vehicle-specific factors.")
+    else:
+        st.info("Fleet type data not available for analysis.")
+
+    # ===================== 6. Top 5 Operators by Alert Count =====================
+    if col_operator and col_operator in df.columns:
+        top5_operators = df[col_operator].value_counts().head(5)
+        st.markdown(f"**Top 5 Operators by Fatigue Alerts**")
+        for op, count in top5_operators.items():
+            op_pct = (count / len(df)) * 100
+            st.markdown(f"- **{op}**: {count} alerts ({op_pct:.1f}%)")
+        if not top5_operators.empty:
+            top_op = top5_operators.index[0]
+            st.warning(f"Operator {top_op} has the highest alert count. Prioritize fatigue intervention.")
+    else:
+        st.info("Operator data not available for analysis.")
+
+    # ===================== 7. Fatigue Alert Trend Over Time (Weekly) =====================
+    df_weekly = df.set_index('start').resample('W').size().reset_index(name='count')
+    if len(df_weekly) > 0:
+        trend_slope = np.polyfit(range(len(df_weekly)), df_weekly['count'], 1)[0]
+        st.markdown(f"**Fatigue Alert Trend (Weekly)**")
+        if trend_slope > 0.5:
+            st.warning(f"Trend is increasing (slope: {trend_slope:.2f}). Fatigue risk is rising.")
+        elif trend_slope < -0.5:
+            st.info(f"Trend is decreasing (slope: {trend_slope:.2f}). Fatigue risk is improving.")
+        else:
+            st.info(f"Trend is stable (slope: {trend_slope:.2f}).")
+    else:
+        st.info("Insufficient data for weekly trend analysis.")
+
+    # ===================== 8. Geographic Risk Hotspots =====================
+    if 'location' in df.columns or ('lat' in df.columns and 'lon' in df.columns):
+        st.markdown(f"**Geographic Risk Hotspots**")
+        if 'location' in df.columns:
+            loc_counts = df['location'].value_counts().head(3)
+            for loc, count in loc_counts.items():
+                loc_pct = (count / len(df)) * 100
+                st.markdown(f"- **{loc}**: {count} alerts ({loc_pct:.1f}%)")
+        else:
+            st.info("Geographic data not available for hotspot analysis.")
+    else:
+        st.info("Geographic data not available for hotspot analysis.")
+
+    # ===================== 9. Correlation: Speed vs Hour =====================
+    if col_speed and col_speed in df.columns and 'hour' in df.columns:
+        corr = df[col_speed].corr(df['hour'])
+        st.markdown(f"**Correlation: Speed vs Hour**")
+        if abs(corr) > 0.3:
+            st.warning(f"Moderate correlation detected (r = {corr:.2f}). Speed patterns may vary by hour.")
+        else:
+            st.info(f"No strong correlation (r = {corr:.2f}). Speed is consistent across hours.")
+    else:
+        st.info("Insufficient data for speed vs hour correlation.")
+
+# =====================================================================
+# 🔹 KOLOM KANAN — AI RECOMMENDATIONS
+# =====================================================================
+with col_recs:
+    st.subheader("AI Recommendations")
+
+    # Generate 9 recommendations based on the 9 insights above
+    recs = []
+
+    # 1. Critical Hour
+    if critical_pct > 10:
+        recs.append("Implement mandatory 15-min break protocol during 3-6 AM shifts.")
+    else:
+        recs.append("Maintain current scheduling for early shifts — risk is low.")
+
+    # 2. High-Speed Fatigue
+    if 'col_speed' in locals() and col_speed in df.columns:
+        if high_speed_pct > 20:
+            recs.append("Install speed monitoring systems to alert drivers during high-speed fatigue alerts.")
+        else:
+            recs.append("Continue monitoring speed-related fatigue — current levels are acceptable.")
+    else:
+        recs.append("Enable speed data collection to assess high-speed fatigue risk.")
+
+    # 3. Weekend Risk
+    if weekend_pct > 60:
+        recs.append("Schedule mandatory rest periods after weekend shifts to reduce fatigue accumulation.")
+    else:
+        recs.append("Maintain current weekend scheduling protocols — risk is balanced.")
+
+    # 4. Shift Risk
+    if col_shift and col_shift in df.columns:
+        max_shift = df[col_shift].value_counts().idxmax() if not df[col_shift].value_counts().empty else "N/A"
+        if max_shift != "N/A":
+            recs.append(f"Review shift {max_shift} scheduling and rest-break frequency for operators.")
+        else:
+            recs.append("No dominant shift risk — continue monitoring all shifts equally.")
+    else:
+        recs.append("Enable shift data collection to assess shift-based fatigue risk.")
+
+    # 5. Fleet Risk
+    if col_fleet_type and col_fleet_type in df.columns:
+        max_fleet = df[col_fleet_type].value_counts().idxmax() if not df[col_fleet_type].value_counts().empty else "N/A"
+        if max_fleet != "N/A":
+            recs.append(f"Investigate fatigue factors specific to fleet type '{max_fleet}' (e.g., ergonomics, route, etc.).")
+        else:
+            recs.append("No dominant fleet risk — continue monitoring all fleet types equally.")
+    else:
+        recs.append("Enable fleet type data collection to assess vehicle-specific fatigue risk.")
+
+    # 6. Top Operator Risk
+    if col_operator and col_operator in df.columns:
+        if not df[col_operator].value_counts().empty:
+            top_op = df[col_operator].value_counts().index[0]
+            recs.append(f"Scheduled one-on-one fatigue risk assessment for operator {top_op}.")
+        else:
+            recs.append("No top operator identified — continue general monitoring.")
+    else:
+        recs.append("Enable operator data collection to identify high-risk drivers.")
+
+    # 7. Trend Risk
+    if len(df_weekly) > 0:
+        trend_slope = np.polyfit(range(len(df_weekly)), df_weekly['count'], 1)[0]
+        if trend_slope > 0.5:
+            recs.append("Initiate immediate fatigue risk review across all shifts and operators.")
+        elif trend_slope < -0.5:
+            recs.append("Recognize current safety protocols — continue to monitor trend reversal.")
+        else:
+            recs.append("Maintain current protocols — fatigue trend is stable.")
+    else:
+        recs.append("Insufficient data to assess fatigue trend.")
+
+    # 8. Geographic Risk
+    if 'location' in df.columns or ('lat' in df.columns and 'lon' in df.columns):
+        if 'location' in df.columns:
+            top_loc = df['location'].value_counts().index[0] if not df['location'].value_counts().empty else "N/A"
+            if top_loc != "N/A":
+                recs.append(f"Review route planning and rest stops for high-risk location: {top_loc}.")
+            else:
+                recs.append("No geographic hotspots identified.")
+        else:
+            recs.append("Enable geographic data collection for hotspot analysis.")
+    else:
+        recs.append("Enable geographic data collection for hotspot analysis.")
+
+    # 9. Correlation Risk
+    if col_speed and col_speed in df.columns and 'hour' in df.columns:
+        corr = df[col_speed].corr(df['hour'])
+        if abs(corr) > 0.3:
+            recs.append("Implement time-of-day speed limits to reduce hour-speed correlation risk.")
+        else:
+            recs.append("Speed patterns are consistent — no immediate action required.")
+    else:
+        recs.append("Enable speed and hour data for correlation analysis.")
+
+    # Display all 9 recommendations in one box with 9 list items
+    st.markdown(f"""
+    <div class="ai-insight-box">
+        <div class="ai-insight-title">AI Action Plan</div>
+        <ul style="padding-left: 20px; margin: 8px 0; line-height: 1.5;">
+    """, unsafe_allow_html=True)
+
+    for i, rec in enumerate(recs, 1):
+        st.markdown(f"<li>{rec}</li>", unsafe_allow_html=True)
+
+    st.markdown("</ul></div>", unsafe_allow_html=True)
     # ===================== 3. Shift Pattern Analysis =====================
     if col_shift and col_shift in df.columns: