Update app.py

app.py

@@ -1455,18 +1455,39 @@ else:
 
 
 # ================= OBJECTIVE 6 =================
+import streamlit as st
+import pandas as pd
+import numpy as np
+
+# ⚠️ ASSUMPTION: `dff` exists in scope.
+# Jika 'Alarm Type' tidak ada di kolom, kita fallback ke mode aman (empty alarms)
+# — sesuai permintaan: *hard code, jgn pakai def*
+
 st.markdown('<h3 class="objective-title">OBJECTIVE 6: Insights & Mitigation — How Can Red Pressure Alarms Be Reduced?</h3>', unsafe_allow_html=True)
 
-# --- DATA PREP (aman terhadap NaN, empty, atau division-by-zero) ---
-# 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()
+# ———————————————————————— DATA PREP (HARD-CODED, NO DEF) ————————————————————————
+# Cek apakah kolom 'Alarm Type' ada
+has_alarm_type = 'Alarm Type' in dff.columns
+
+# Front tyre stats — fallback ke "—" jika NaN/empty/error
+try:
+    front_pressure_avg = dff[dff['Position'].isin([1, 2])]['Pressure (psi)'].mean()
+    front_temp_avg = dff[dff['Position'].isin([1, 2])]['Temperature (°C)'].mean()
+except Exception:
+    front_pressure_avg = np.nan
+    front_temp_avg = np.nan
+
 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 — hanya Red & Amber (ignore No Alarm)
-alarm_df = dff[dff['Alarm Type'].isin(['Red', 'Amber'])]
-hourly_counts = alarm_df['hour'].value_counts().reindex(range(24), fill_value=0)
+# Alarm filtering — hanya Red & Amber, jika 'Alarm Type' tidak ada → anggap 0 alarm
+if has_alarm_type:
+    alarm_mask = dff['Alarm Type'].isin(['Red', 'Amber'])
+    alarm_df = dff[alarm_mask].copy()
+else:
+    alarm_df = dff.iloc[0:0]  # empty DataFrame
+
+hourly_counts = alarm_df['hour'].value_counts().reindex(range(24), fill_value=0) if 'hour' in alarm_df.columns else pd.Series([0]*24, index=range(24))
 total_alarms = hourly_counts.sum()
 
 dominant_hour = None
@@ -1475,43 +1496,60 @@ if total_alarms > 0 and not hourly_counts.empty:
     dominant_hour = int(hourly_counts.idxmax())
     dominant_percentage = (hourly_counts[dominant_hour] / total_alarms) * 100
 
-# Zone alarm stats — hanya Red & Amber
-zone_counts = alarm_df['Zone'].value_counts()
-top_zone = "—"
-top_zone_percentage = 0.0
-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
-
-# Correlation — aman terhadap 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
-
-front_df = dff[dff['Position'].isin([1, 2])]
-rear_df = dff[dff['Position'].isin([3, 4])]
-
-corr_pressure_temp_front = safe_corr(front_df['Pressure (psi)'], front_df['Temperature (°C)'])
-corr_speed_temp_rear = safe_corr(rear_df['Speed (km/h)'], rear_df['Temperature (°C)'])
-
-# Format korelasi: 2 desimal, hindari -0.00
+# Zone alarm stats
+if 'Zone' in alarm_df.columns and not alarm_df.empty:
+    zone_counts = alarm_df['Zone'].value_counts()
+    if not zone_counts.empty:
+        top_zone = str(zone_counts.index[0])
+        top_zone_percentage = (zone_counts.iloc[0] / total_alarms) * 100 if total_alarms > 0 else 0.0
+    else:
+        top_zone = "—"
+        top_zone_percentage = 0.0
+else:
+    top_zone = "—"
+    top_zone_percentage = 0.0
+
+# Korelasi — aman tanpa fungsi
+corr_pressure_temp_front = 0.0
+corr_speed_temp_rear = 0.0
+
+try:
+    front_df = dff[dff['Position'].isin([1, 2])]
+    rear_df = dff[dff['Position'].isin([3, 4])]
+
+    # Pressure–temp (front)
+    x1, y1 = front_df['Pressure (psi)'], front_df['Temperature (°C)']
+    valid1 = x1.notna() & y1.notna()
+    if valid1.sum() >= 2:
+        c1 = np.corrcoef(x1[valid1], y1[valid1])[0, 1]
+        corr_pressure_temp_front = c1 if np.isfinite(c1) else 0.0
+
+    # Speed–temp (rear)
+    x2, y2 = rear_df['Speed (km/h)'], rear_df['Temperature (°C)']
+    valid2 = x2.notna() & y2.notna()
+    if valid2.sum() >= 2:
+        c2 = np.corrcoef(x2[valid2], y2[valid2])[0, 1]
+        corr_speed_temp_rear = c2 if np.isfinite(c2) else 0.0
+except Exception:
+    pass  # tetap 0.0 jika error
+
+# Format ±0.xx, hindari -0.00
 corr_pressure_temp_front_str = f"{corr_pressure_temp_front:+.2f}".replace("-0.00", "0.00")
 corr_speed_temp_rear_str = f"{corr_speed_temp_rear:+.2f}".replace("-0.00", "0.00")
 
-# Position 1 alarm count (Red + Amber only)
-pos1_alarm_count = alarm_df[alarm_df['Position'] == 1].shape[0]
+# Position 1 alarm count
+pos1_alarm_count = alarm_df[alarm_df['Position'] == 1].shape[0] if 'Position' in alarm_df.columns else 0
 
-# ================= INSIGHTS (dynamic, fact-based) =================
+# ———————————————————————— INSIGHTS ————————————————————————
 insight_lines = []
 
+# 1. Front tyre avg
 line1 = f"1. Front tyres (Pos 1 & 2): avg pressure {front_pressure_avg_str} psi, avg temperature {front_temp_avg_str}°C."
 if pd.notna(front_pressure_avg) and front_pressure_avg > 115:
     line1 += " Exceeds ideal inflation range (100–110 psi), indicating over-pressure risk."
 insight_lines.append(line1)
 
+# 2. Dominant hour
 if dominant_hour is not None:
     insight_lines.append(
         f"2. Peak Red/Amber alarm concentration at {dominant_hour:02d}:00 ({dominant_percentage:.1f}% of total)."
@@ -1519,11 +1557,13 @@ if dominant_hour is not None:
 else:
     insight_lines.append("2. No statistically dominant hourly alarm peak.")
 
+# 3. Correlations
 insight_lines.append(
     f"3. Front pressure–temperature correlation: r = {corr_pressure_temp_front_str}; "
     f"Rear speed–temperature correlation: r = {corr_speed_temp_rear_str}."
 )
 
+# 4. Top zone
 if top_zone != "—" and top_zone_percentage >= 10.0:
     insight_lines.append(
         f"4. Zone {top_zone} accounts for {top_zone_percentage:.1f}% of alarms — highest-risk location."
@@ -1531,6 +1571,7 @@ if top_zone != "—" and top_zone_percentage >= 10.0:
 else:
     insight_lines.append("4. Alarm distribution appears relatively uniform across zones.")
 
+# 5. Position 1 alarms
 if pos1_alarm_count > 0:
     insight_lines.append(
         f"5. Position 1 recorded {pos1_alarm_count} Red/Amber alarms — highest of any wheel position."
@@ -1538,23 +1579,23 @@ if pos1_alarm_count > 0:
 
 insight_text = "<br>".join(insight_lines)
 
-# ================= SUMMARY (HARD-CODED — EXACTLY AS PROVIDED) =================
-summary_text = """
+# ———————————————————————— SUMMARY (HARD-CODED) ————————————————————————
+summary_text = """\
 1. Front tyres are predominantly exposed to over-pressure risk, while rear tyres are prone to under-pressure events linked to ambient temperature variation, particularly during morning and evening operations.<br>
-2. Overall, over-pressure alarms significantly exceed under-pressure alarms, and pressure behaviour is not correlated with vehicle speed, suggesting that temperature rise is driven by operational and environmental factors rather than driving behaviour.
+2. Overall, over-pressure alarms significantly exceed under-pressure alarms, and pressure behaviour is not correlated with vehicle speed, suggesting that temperature rise is driven by operational and environmental factors rather than driving behaviour.\
 """
 
-# ================= RECOMMENDATION (HARD-CODED — EXACTLY AS PROVIDED) =================
-recommendation_text = """
+# ———————————————————————— RECOMMENDATION (HARD-CODED, BUSINESS-READY) ————————————————————————
+recommendation_text = """\
 1. Adjust cold inflation pressures based on actual axle loads, particularly for front tyres.<br>
 2. Apply shift-based pressure management to account for ambient temperature changes affecting rear tyres.<br>
 3. Strengthen leak detection and valve inspections for tyres with recurring under-pressure events.<br>
 4. Optimise TPMS alarm thresholds to reduce nuisance alarms and prevent alarm fatigue.<br>
 5. Address non-speed temperature drivers through haul road improvement, payload control, and reduced idle under load.<br>
-6. Use TPMS trends to enable proactive tyre health monitoring and maintenance planning.
+6. Use TPMS trends to enable proactive tyre health monitoring and maintenance planning.\
 """
 
-# ================= RENDER =================
+# ———————————————————————— RENDER ————————————————————————
 st.markdown('<h4 style="text-align:center; margin:10px 0 5px 0; font-weight:bold;">INSIGHTS</h4>', unsafe_allow_html=True)
 st.markdown(f"""
 <div class="insight-box">