API_README.md

# Castor Price Forecasting API

A REST API for accessing ARIMA and LSTM forecasting models for Castor oil prices.

## Getting Started

### 1. Generate an API Key

```bash
curl -X POST http://localhost:5000/api/keys/generate \
  -H "Content-Type: application/json" \
  -d '{"name": "my-app"}'
```

**Response:**
```json
{
  "status": "success",
  "api_key": "castor_api_a1b2c3d4e5f6g7h8i9j0k1l2m3n4o5p6",
  "message": "API key generated: castor_api_a1b2c3d4e5f6g7h8i9j0k1l2m3n4o5p6"
}
```

### 2. Use Your API Key

Include the API key in the `X-API-Key` header for all requests:

```bash
curl -X POST http://localhost:5000/api/forecast/arima \
  -H "X-API-Key: castor_api_a1b2c3d4e5f6g7h8i9j0k1l2m3n4o5p6" \
  -H "Content-Type: application/json" \
  -d '{
    "product": "Castor",
    "start_date": "2025-12-01",
    "end_date": "2026-12-31"
  }'
```

## API Endpoints

### Health Check
```
GET /api/health
```
Check if the API is running.

### Key Management

#### Generate API Key
```
POST /api/keys/generate
Body: {"name": "my-app"}
```

#### List All Keys
```
GET /api/keys/list
```

#### Revoke API Key
```
POST /api/keys/revoke
Body: {"api_key": "castor_api_..."}
```

### Forecasting

#### ARIMA Forecast
```
POST /api/forecast/arima
Headers: X-API-Key: YOUR_KEY
Body: {
  "product": "Castor",
  "start_date": "2025-12-01",
  "end_date": "2026-12-31"
}
```

#### LSTM Forecast
```
POST /api/forecast/lstm
Headers: X-API-Key: YOUR_KEY
Body: {
  "product": "Castor",
  "start_date": "2025-12-01",
  "end_date": "2026-12-31"
}
```

#### Compare Models
```
POST /api/forecast/compare
Headers: X-API-Key: YOUR_KEY
Body: {
  "product": "Castor",
  "start_date": "2025-12-01",
  "end_date": "2026-12-31"
}
```

## Example Usage (Python)

```python
import requests

# Generate API Key
response = requests.post('http://localhost:5000/api/keys/generate', 
  json={'name': 'my-app'})
api_key = response.json()['api_key']

# Get forecast
headers = {'X-API-Key': api_key}
payload = {
  'product': 'Castor',
  'start_date': '2025-12-01',
  'end_date': '2026-12-31'
}

forecast = requests.post('http://localhost:5000/api/forecast/arima',
  json=payload, headers=headers)
print(forecast.json())
```

## Running the API Server

```bash
# Using the virtual environment
D:\models\arima\venv_short\Scripts\python.exe api_server.py
```

The server will start at `http://localhost:5000`

## API Response Format

### Success Response
```json
{
  "status": "success",
  "product": "Castor",
  "model": "ARIMA",
  "forecast": {
    "2025-12-01": 3500.5,
    "2025-12-02": 3502.3,
    ...
  },
  "start_date": "2025-12-01",
  "end_date": "2026-12-31"
}
```

### Error Response
```json
{
  "status": "error",
  "message": "Invalid or missing API key"
}
```

## API Key Security

- Each API key is tracked with:
  - Creation timestamp
  - Last used timestamp
  - Request count
  - Custom name/identifier

- Keys are stored in `api_keys.json`
- In production, use a proper database with encryption
- Revoke keys when no longer needed

## Features

✅ API Key-based authentication  
✅ ARIMA forecasting  
✅ LSTM forecasting  
✅ Model comparison  
✅ Request tracking  
✅ Multiple products support  
✅ Flexible date ranges  
✅ Cross-Origin Resource Sharing (CORS)  

## Troubleshooting

**"Invalid or missing API key"**: Ensure you're including the correct API key in the `X-API-Key` header.

**"Product not found"**: Check that the product name matches the data in your CSV file.

**Server not responding**: Make sure the Flask server is running: `python api_server.py`