CA Delivery Van Metrics Analysis

When to Use

• "Analyze this week's metrics"
• "Check weekly performance"
• "Validate dashboard data"
• "Review van performance"
• "How are the carriers doing?"
• "Check CPD trends"
• User asks about FOX/NTG/FDC performance

Quick Analysis

For immediate results:

# Load and analyze data
python3 << 'EOF'
from scripts.analysis.weekly_metrics import analyze_weekly_metrics
from scripts.analysis.cpd_analysis import load_rate_cards
import pandas as pd

nash_df = pd.read_csv('data/nash_data.csv')
rate_cards = load_rate_cards('data/rate_cards.json')

results = analyze_weekly_metrics(nash_df, rate_cards, min_batch_size=10)

print(f"📊 Latest Week: {results['weeks'][-1]['week_start']}")
print(f"   Orders: {results['weeks'][-1]['total_orders']}")
print(f"   CPD: ${results['weeks'][-1]['weighted_avg_cpd']:.2f}")
print(f"   Excluded: {results['weeks'][-1]['excluded_count']} trips")
EOF

Complete Workflow

Step 1: Data Validation

python3 -c "
import pandas as pd
df = pd.read_csv('data/nash_data.csv')
print(f'✅ Loaded {len(df)} trips')
print(f'📅 {df[\"Date\"].min()} to {df[\"Date\"].max()}')
print(f'🏪 {df[\"Store Id\"].nunique()} stores')
print(f'🚚 Carriers: {list(df[\"Carrier\"].unique())}')
"

Validation Checklist:

• Nash CSV loaded
• Full weeks (Monday-Sunday)
• 273 CA stores
• 3 carriers (FOX, NTG, FDC)

Step 2: Calculate Metrics

Run weekly analysis with anomaly exclusion (batches < 10 orders):

from scripts.analysis.weekly_metrics import analyze_weekly_metrics
from scripts.analysis.cpd_analysis import load_rate_cards
import pandas as pd

nash_df = pd.read_csv('data/nash_data.csv')
rate_cards = load_rate_cards('data/rate_cards.json')

# min_batch_size=10 excludes anomalies
results = analyze_weekly_metrics(nash_df, rate_cards, min_batch_size=10)

Metrics Calculated:

• Total orders/trips/batches per week
• Weighted average CPD (anomalies excluded)
• Carrier performance breakdown
• Store performance rankings
• Exclusion statistics

Step 3: Analyze Trends

Key Questions:

1. CPD Trending

   • Compare last 2 weeks
   • Check if within $380-$400 range
   • Flag if change > 5%

2. Carrier Performance

   • FOX should be primary (highest volume)
   • Check for unusual CPD spikes
   • Validate carrier mix

3. Anomaly Exclusions

   • Should be < 5% of total trips
   • All excluded trips < 10 orders

4. Store Patterns

   • Top 10 stores consistent week-over-week
   • No single store > 30% volume

Step 4: Dashboard Validation

Verify dashboard matches Python calculations:

npm run test:integration

Checks:

• Weekly chart CPD aligns
• Carrier comparison uses same exclusion logic
• OTD percentages match Nash data
• Batch density excludes anomalies
• CSV export has all fields

Step 5: Generate Report

Standard Format:

📊 CA DELIVERY VANS - WEEKLY METRICS
Week: [Start] to [End]

📈 KEY METRICS:
- Orders: [X] (+/- Y% vs last week)
- CPD: $[X.XX] (+/- $Y.YY vs last week)
- Trips: [X]
- Excluded: [X] trips (Z%)

🚚 CARRIER PERFORMANCE:
- FOX: [X] trips, $[X.XX] CPD (Primary)
- NTG: [X] trips, $[X.XX] CPD
- FDC: [X] trips, $[X.XX] CPD

⚠️ ISSUES:
[List concerns or "None"]

✅ VALIDATION:
- Data quality: Pass/Fail
- Anomaly exclusion: Pass/Fail
- Dashboard alignment: Pass/Fail

Validation Gates

All must pass:

# Data integrity
python3 -m pytest tests/unit/analysis_test.py

# Integration
npm run test:integration

# Dashboard rendering
# Open dashboard.html, verify charts render

Common Issues

CPD Mismatch

Symptom: Dashboard ≠ Python Fix: Check analytics.service.ts uses min_batch_size=10

Missing Weeks

Symptom: Chart gaps Fix: Verify get_week_start() returns Monday

High Exclusions

Symptom: > 10% excluded Fix: Verify threshold is 10 (not higher)

Carrier Names

Symptom: Duplicate carriers Fix: Use normalize_carrier_name()

Key Files

Python:

• scripts/analysis/weekly_metrics.py - Core logic
• scripts/analysis/cpd_analysis.py - CPD calculation
• scripts/analysis/dashboard.py - Chart data

Frontend:

• public/weekly-metrics.html - Weekly chart
• public/js/weekly-metrics.js - Rendering
• public/dashboard.html - Main dashboard

Data:

• data/nash_data.csv - Trip data
• data/rate_cards.json - Carrier rates
• States/walmart_stores_ca_only.csv - Store mapping

Tests:

• tests/unit/analysis_test.py - Python tests
• tests/integration/dashboard-flow.test.ts - E2E tests

Output Example

✅ Analyzing CA Delivery Van Metrics

📊 DATA VALIDATION
✅ Loaded 1,247 trips from nash_data.csv
✅ Date range: 2025-10-07 to 2025-10-13
✅ 273 CA stores detected
✅ All 3 carriers present

📈 WEEKLY METRICS (Oct 7-13, 2025)
- Total Orders: 12,458
- Average CPD: $387.45
- Total Trips: 1,247
- Excluded: 23 trips (1.8%)

🚚 CARRIER PERFORMANCE
- FOX: 856 trips, $385.20 CPD (68.6%)
- NTG: 298 trips, $392.10 CPD (23.9%)
- FDC: 93 trips, $391.80 CPD (7.5%)

📊 WEEK-OVER-WEEK
- CPD: $387.45 vs $389.20 (-$1.75, -0.4%) ✅
- Orders: 12,458 vs 11,923 (+535, +4.5%)

⚠️ ISSUES: None

✅ VALIDATION: All checks passed

📊 Dashboard: http://localhost:3000/weekly-metrics

Progressive Disclosure

For deep-dives, see:

• reference/cpd-calculation.md - CPD formula details
• reference/anomaly-detection.md - Exclusion logic
• reference/carrier-analysis.md - Vendor performance metrics

Version: 1.0.0 Created: 2025-10-17 Type: Official Agent Skill (Anthropic format)