ipm-advisor

# ipm-advisor

Analyze pest pressure data from pest-disease-log to identify intervention needs, evaluate spray effectiveness, and guide IPM strategy decisions.

## When to Use

- Identifying which fields have pressure above action thresholds
- Comparing pressure before and after a spray application
- Ranking pests by frequency and severity across the season
- Estimating whether spray programs are cost-effective
- Finding fields where pressure is increasing (trend analysis)

## Action Threshold Logic

The pest_library table stores an `action_threshold` integer (0-10) per pest. When a field's current severity for a pest exceeds this threshold, intervention is recommended.

```bash
# Get pest library with action thresholds
curl http://localhost:3006/api/pest-library

# Example: Aphids action threshold = 6
# If current observation severity = 8, action is overdue
```

## Identifying Fields Needing Intervention

```bash
# Get field summary (latest severity per field)
curl http://localhost:3006/api/analytics/field-summary

# Response includes per-field max_severity and latest pest observations
# Filter for fields where max_severity >= action threshold for any active pest
```

Fields with `max_severity >= 7` (High or Critical) generally require immediate action.

## Evaluating Spray Effectiveness

To compare pressure before and after a spray:

```bash
# 1. Get the spray record date
curl "http://localhost:3006/api/sprays?field_id={id}"

# 2. Get observations before the spray (last 2 weeks prior)
curl "http://localhost:3006/api/observations?field_id={id}&to={spray_date}&pest_type=Aphids"

# 3. Get observations after the spray (next 2 weeks)
curl "http://localhost:3006/api/observations?field_id={id}&from={spray_date}&pest_type=Aphids"

# Compare average severity before vs. after
# Effective spray: post-spray avg severity < pre-spray avg severity by >= 2 points
```

## Pressure Trend Analysis

```bash
# Get severity trend over time for all pests in a field
curl "http://localhost:3006/api/analytics/pressure-trend?field_id={id}&from=2025-06-01"

# Response: [{ date, pest_type, avg_severity }]
# Increasing trend: last 3 observations each higher than previous
# Decreasing trend: last 3 observations each lower than previous
```

A pest is "trending up" if three consecutive observations show increasing severity. This is an early warning to act before reaching the action threshold.

## Season Ranking

```bash
# Most frequent pests this season
curl "http://localhost:3006/api/analytics/pest-frequency?from=2025-04-01"

# IPM summary: per-field obs count, avg severity, spray count, cost
curl http://localhost:3006/api/analytics/ipm-summary
```

From the IPM summary, calculate:
- **Spray efficiency**: `total_cost / observations_count` - lower is better
- **Pressure-to-spray ratio**: `avg_severity / spray_count` - high ratio means under-treated
- **Field priority**: sort fields by `max_severity DESC, last_scouted ASC`

## Spray Cost Analysis

```bash
# Spray cost by field and month
curl "http://localhost:3006/api/analytics/spray-cost?from=2025-04-01"

# Per-field cost breakdown
curl "http://localhost:3006/api/sprays?field_id={id}&from=2025-04-01"
```

## IPM Recommendations

| Situation | Recommendation |
|---|---|
| Severity >= action threshold | Apply targeted treatment |
| Severity trending up for 3+ observations | Scout more frequently, prepare treatment |
| High severity but spray applied recently | Wait for PHI to clear, re-scout |
| Low severity after spray | Treatment effective, continue monitoring |
| No spray applied but severity >= 7 | Immediate action needed |
| Severity 1-3, stable | Monitor weekly |

## Before/After Report Pattern

To generate a before/after report for an amendment:

1. Get all spray records for a field: `GET /api/sprays?field_id={id}`
2. For each spray, find observations in 14 days before: `GET /api/observations?field_id={id}&pest_type={pest}&to={spray_date}&limit=5`
3. Find observations in 14 days after: `GET /api/observations?field_id={id}&pest_type={pest}&from={spray_date}&limit=5`
4. Calculate: `before_avg = mean(pre_severities)`, `after_avg = mean(post_severities)`, `reduction = before_avg - after_avg`
5. If `reduction >= 2.0`: treatment was effective. If reduction < 1.0: treatment had minimal effect.

## Common IPM Questions

### Which field has the worst pressure right now?

```bash
curl http://localhost:3006/api/analytics/field-summary
# Sort by max_severity DESC, take the top result
```

### Which pest is hardest to control this season?

```bash
# Get observations where severity stayed >= 6 even after sprays
# Cross-reference pest_frequency with spray records per pest_type
```

### Is pressure increasing or decreasing on North Field?

```bash
curl "http://localhost:3006/api/analytics/pressure-trend?field_id={north_field_id}&from=2025-07-01"
# Look at the last 3 data points for each pest_type
# Compare first avg_severity to last avg_severity
```