alpine-agent / src /tools.py
florentgbelidji's picture
Added source modules of the app
7f4e4c3
raw
history blame
7.51 kB
import os
import pandas as pd
from smolagents import Tool
from typing import List, Dict, Any, Union, Tuple
from meteofrance_api import MeteoFranceClient
from src.skitour_api import get_topos, get_refuges, get_details_topo, get_massifs, get_recent_outings
from src.meteo_france_api import get_massif_conditions
from src.utils import geocode_location, assign_location_to_clusters, haversine, llm_summarizer
class RefugeTool(Tool):
name = "refuge_recherche"
description = "Recherche d'un refuge dans un massif donné"
inputs = {
"massif_id": {
"description": "[Optional, default: None] Id du massif souhaité ",
"type": "string",
}
}
output_type = "string"
def forward(self, massif_id) -> List[Dict]:
return get_refuges(massif_id)
class GetRoutesTool(Tool):
name = "list_routes"
description = """
Looking for a list of ski touring routes in a given list of mountain ranges.
Returns a list containing the information of the topos found.
Use `topo_details` immediately after this tool to get the details of a specific topo.
"""
inputs = {
"mountain_range_ids": {
"description": "List of mountain range ids",
"type": "string",
}
}
output_type = "any"
def forward(self, mountain_range_ids: str) -> List[Dict]:
topos = get_topos(mountain_range_ids)
return topos
class DescribeRouteTool(Tool):
name = "describe_route"
description = """
Searches for key information about a specific ski touring route, including weather forecasts and associated avalanche risks.
Always use this tool after using the `list_routes` tool.
This tool returns a dictionary containing the route's information, the avalanche risk estimation bulletin, and the weather forecast for the coming days of the route.
"""
inputs = {
"id_route": {
"description": "id of the route",
"type": "string",
},
"id_range": {
"description": "mountain range id of the route",
"type": "string"}
}
output_type = "any"
def __init__(self, skitour2meteofrance: dict, llm_engine: Any):
super().__init__()
self.massifs_infos = skitour2meteofrance
self.weather_client = MeteoFranceClient(access_token=os.getenv("METEO_FRANCE_API_KEY"))
self.llm_engine = llm_engine
def forward(self, id_route: str, id_range: str) -> dict:
topo_info = get_details_topo(str(id_route))
avalanche_conditions = get_massif_conditions(
self.massifs_infos[str(id_range)]['meteofrance_id']
)
lat, lon = topo_info["depart"]["latlon"]
weather_forecast = self.weather_client.get_forecast(float(lat), float(lon))
daily_forecast = weather_forecast.forecast[:24]
for day_forecast in daily_forecast:
day_forecast["dt"] = weather_forecast.timestamp_to_locale_time(day_forecast["dt"]).isoformat()
forecast_summary = llm_summarizer(str(daily_forecast), self.llm_engine)
avalanche_summary = llm_summarizer(str(avalanche_conditions), self.llm_engine)
return {
"route_info": topo_info,
"avalanche_conditions": avalanche_summary,
"daily_weather_forecast": forecast_summary,
"route_link": f"https://skitour.fr/topos/{id_route}"
}
class RecentOutingsTool(Tool):
name = "recent_outings"
description = """
Searches for recent outings in a given mountain range.
Returns a list of the most recent outings in the given range.
"""
inputs = {
"id_range": {
"description": "id of the mountain range",
"type": "string",
}
}
output_type = "any"
def forward(self, id_range: str) -> List[Dict]:
return get_recent_outings(id_range)
class MountainRangesTool(Tool):
name = "list_mountain_ranges"
description = """ Searches for the ID(s) of the mountain ranges closest to a given location.
If the location is too far from known ranges, the search returns None.
Should return a string with the massif IDs separated by commas.
"""
inputs = {
"location": {
"description": "Location to search for",
"type": "string",
},
"num_ranges": {
"description": "[Optional, default: 3] Number of closest mountain ranges to return",
"type": "number",
}
}
output_type = "string"
def __init__(self, clusters: Dict[str, List[Tuple[float, float]]]):
super().__init__()
self.clusters = clusters
def forward(self, location: str, num_ranges: int) -> Union[str, None]:
coord_location = geocode_location(location)
if not location:
return None
matched_ranges = assign_location_to_clusters(coord_location, self.clusters, k=num_ranges)
list_ranges = [range[0] for range in matched_ranges if range[1] < 100]
if not list_ranges:
return ''
massifs= get_massifs()
massif_ids = [_massif['id'] for _massif in massifs if _massif['nom'] in list_ranges]
return ", ".join(massif_ids)
class ForecastTool(Tool):
name = "forecast"
description = """Searches for the weather forecast for a given location as well as the current avalanche risk estimation bulletin.
Unnecessary if the user is inquiring about a route, as `describe_route` already provides this information."""
inputs = {
"location": {
"description": "Location to search for",
"type": "string",
},
}
output_type = "any"
def __init__(self, llm_engine, clusters: Dict[str, List[Tuple[float, float]]], skitour2meteofrance: dict):
super().__init__()
self.clusters = clusters
self.massifs_infos = skitour2meteofrance
self.llm_engine = llm_engine
def forward(self, location: str) -> Union[Dict[str, Any], None]:
coord_location = geocode_location(location)
if not location:
return None
# Get the closest mountain range to the location to get the avalanche conditions
matched_ranges = assign_location_to_clusters(coord_location, self.clusters, k=1)
list_ranges = [range[0] for range in matched_ranges if range[1] < 100]
if not list_ranges:
return None
massifs= get_massifs()
massif_id = [_massif['id'] for _massif in massifs if _massif['nom'] in list_ranges]
avalanche_conditions = get_massif_conditions(
self.massifs_infos[str(massif_id[0])]['meteofrance_id']
)
weather_client = MeteoFranceClient(access_token=os.getenv("METEO_FRANCE_API_KEY"))
forecast = weather_client.get_forecast(*coord_location)
daily_forecast = forecast.forecast[:24]
for day_forecast in daily_forecast:
day_forecast["dt"] = forecast.timestamp_to_locale_time(day_forecast["dt"]).isoformat()
forecast_summary = llm_summarizer(str(daily_forecast), self.llm_engine)
avalanche_summary = llm_summarizer(str(avalanche_conditions), self.llm_engine)
return {"forecast": forecast_summary, "avalanche_conditions": avalanche_summary}