Weather API Agent Skill

import requests
import pandas as pd
from typing import Dict, Any, List, Optional
from dataclasses import dataclass
from datetime import datetime, timedelta
from enum import Enum


class WeatherRisk(Enum):
    """Weather risk levels for construction."""
    LOW = "low"
    MODERATE = "moderate"
    HIGH = "high"
    CRITICAL = "critical"


@dataclass
class WeatherCondition:
    """Weather condition at a point in time."""
    timestamp: datetime
    temperature: float  # Celsius
    humidity: float     # Percent
    wind_speed: float   # m/s
    precipitation: float  # mm
    conditions: str


@dataclass
class WorkabilityAssessment:
    """Assessment of weather workability."""
    date: datetime
    risk_level: WeatherRisk
    workable_hours: int
    affected_activities: List[str]
    recommendations: List[str]


class WeatherAPIClient:
    """Client for weather APIs."""

    # Free tier endpoints
    OPEN_METEO_BASE = "https://api.open-meteo.com/v1"

    def __init__(self, api_key: Optional[str] = None):
        self.api_key = api_key

    def get_forecast(self, latitude: float, longitude: float,
                     days: int = 7) -> List[WeatherCondition]:
        """Get weather forecast."""
        url = f"{self.OPEN_METEO_BASE}/forecast"
        params = {
            'latitude': latitude,
            'longitude': longitude,
            'hourly': 'temperature_2m,relative_humidity_2m,wind_speed_10m,precipitation',
            'forecast_days': days
        }

        response = requests.get(url, params=params)
        if response.status_code != 200:
            raise Exception(f"API error: {response.status_code}")

        data = response.json()
        return self._parse_forecast(data)

    def get_historical(self, latitude: float, longitude: float,
                       start_date: str, end_date: str) -> List[WeatherCondition]:
        """Get historical weather data."""
        url = f"{self.OPEN_METEO_BASE}/archive"
        params = {
            'latitude': latitude,
            'longitude': longitude,
            'start_date': start_date,
            'end_date': end_date,
            'hourly': 'temperature_2m,relative_humidity_2m,wind_speed_10m,precipitation'
        }

        response = requests.get(url, params=params)
        if response.status_code != 200:
            raise Exception(f"API error: {response.status_code}")

        data = response.json()
        return self._parse_forecast(data)

    def _parse_forecast(self, data: Dict) -> List[WeatherCondition]:
        """Parse API response to WeatherCondition list."""
        conditions = []
        hourly = data.get('hourly', {})

        times = hourly.get('time', [])
        temps = hourly.get('temperature_2m', [])
        humidity = hourly.get('relative_humidity_2m', [])
        wind = hourly.get('wind_speed_10m', [])
        precip = hourly.get('precipitation', [])

        for i in range(len(times)):
            conditions.append(WeatherCondition(
                timestamp=datetime.fromisoformat(times[i]),
                temperature=temps[i] if i < len(temps) else 0,
                humidity=humidity[i] if i < len(humidity) else 0,
                wind_speed=wind[i] if i < len(wind) else 0,
                precipitation=precip[i] if i < len(precip) else 0,
                conditions=self._describe_conditions(
                    temps[i] if i < len(temps) else 0,
                    precip[i] if i < len(precip) else 0,
                    wind[i] if i < len(wind) else 0
                )
            ))

        return conditions

    def _describe_conditions(self, temp: float, precip: float, wind: float) -> str:
        """Generate weather description."""
        conditions = []

        if temp < 0:
            conditions.append("Freezing")
        elif temp > 35:
            conditions.append("Extreme heat")
        elif temp > 30:
            conditions.append("Hot")
        elif temp < 10:
            conditions.append("Cold")

        if precip > 10:
            conditions.append("Heavy rain")
        elif precip > 2:
            conditions.append("Rain")
        elif precip > 0:
            conditions.append("Light rain")

        if wind > 15:
            conditions.append("Strong winds")
        elif wind > 10:
            conditions.append("Windy")

        return ", ".join(conditions) if conditions else "Clear"

    def to_dataframe(self, conditions: List[WeatherCondition]) -> pd.DataFrame:
        """Convert conditions to DataFrame."""
        data = [{
            'timestamp': c.timestamp,
            'temperature': c.temperature,
            'humidity': c.humidity,
            'wind_speed': c.wind_speed,
            'precipitation': c.precipitation,
            'conditions': c.conditions
        } for c in conditions]
        return pd.DataFrame(data)


class ConstructionWeatherRisk:
    """Assess weather risk for construction activities."""

    # Activity-specific thresholds
    THRESHOLDS = {
        'concrete_pour': {
            'min_temp': 5, 'max_temp': 35,
            'max_wind': 12, 'max_precip': 0.5
        },
        'crane_work': {
            'min_temp': -10, 'max_temp': 40,
            'max_wind': 10, 'max_precip': 5
        },
        'exterior_paint': {
            'min_temp': 10, 'max_temp': 35,
            'max_wind': 8, 'max_precip': 0
        },
        'roofing': {
            'min_temp': 5, 'max_temp': 38,
            'max_wind': 12, 'max_precip': 0
        },
        'earthwork': {
            'min_temp': -5, 'max_temp': 40,
            'max_wind': 20, 'max_precip': 10
        }
    }

    def assess_workability(self, condition: WeatherCondition,
                           activities: List[str] = None) -> WorkabilityAssessment:
        """Assess workability for given conditions."""

        if activities is None:
            activities = list(self.THRESHOLDS.keys())

        affected = []
        recommendations = []

        for activity in activities:
            if activity in self.THRESHOLDS:
                thresh = self.THRESHOLDS[activity]

                reasons = []
                if condition.temperature < thresh['min_temp']:
                    reasons.append(f"Too cold ({condition.temperature}°C)")
                if condition.temperature > thresh['max_temp']:
                    reasons.append(f"Too hot ({condition.temperature}°C)")
                if condition.wind_speed > thresh['max_wind']:
                    reasons.append(f"High wind ({condition.wind_speed} m/s)")
                if condition.precipitation > thresh['max_precip']:
                    reasons.append(f"Precipitation ({condition.precipitation} mm)")

                if reasons:
                    affected.append(activity)
                    recommendations.append(f"{activity}: " + ", ".join(reasons))

        # Determine overall risk level
        if len(affected) >= len(activities) * 0.8:
            risk = WeatherRisk.CRITICAL
            workable = 0
        elif len(affected) >= len(activities) * 0.5:
            risk = WeatherRisk.HIGH
            workable = 4
        elif len(affected) > 0:
            risk = WeatherRisk.MODERATE
            workable = 6
        else:
            risk = WeatherRisk.LOW
            workable = 8

        return WorkabilityAssessment(
            date=condition.timestamp,
            risk_level=risk,
            workable_hours=workable,
            affected_activities=affected,
            recommendations=recommendations
        )

    def weekly_forecast_risk(self, conditions: List[WeatherCondition],
                             activities: List[str] = None) -> pd.DataFrame:
        """Assess risk for week of weather data."""

        # Group by date
        daily_conditions = {}
        for c in conditions:
            date = c.timestamp.date()
            if date not in daily_conditions:
                daily_conditions[date] = []
            daily_conditions[date].append(c)

        assessments = []
        for date, day_conditions in daily_conditions.items():
            # Use midday condition as representative
            midday = [c for c in day_conditions
                      if 10 <= c.timestamp.hour <= 16]
            representative = midday[len(midday)//2] if midday else day_conditions[0]

            assessment = self.assess_workability(representative, activities)
            assessments.append({
                'date': date,
                'risk_level': assessment.risk_level.value,
                'workable_hours': assessment.workable_hours,
                'affected_count': len(assessment.affected_activities)
            })

        return pd.DataFrame(assessments)

# Initialize client
weather = WeatherAPIClient()

# Get forecast for site
conditions = weather.get_forecast(latitude=52.52, longitude=13.41, days=7)
df = weather.to_dataframe(conditions)
print(df.head())

# Assess construction risk
risk = ConstructionWeatherRisk()
weekly_risk = risk.weekly_forecast_risk(conditions)
print(weekly_risk)

conditions = weather.get_forecast(52.52, 13.41, days=14)
risk = ConstructionWeatherRisk()

# Check concrete pour window
for c in conditions:
    assessment = risk.assess_workability(c, ['concrete_pour'])
    if assessment.risk_level == WeatherRisk.LOW:
        print(f"Good for concrete: {c.timestamp}")

historical = weather.get_historical(52.52, 13.41, '2024-01-01', '2024-03-31')
df = weather.to_dataframe(historical)

# Count rain days
rain_days = df[df['precipitation'] > 2]['timestamp'].dt.date.nunique()
print(f"Rain days in Q1: {rain_days}")