f1-mcp-server / fastf1_tools.py
arre99's picture
fix conversion for round parameter so it is converted to integer if isnumeric is true. Added some extra comments to make it more clear to the user how to interact with the Gradio UI
18c04d1
raw
history blame
10.6 kB
"""
This module provides tools for interacting with Formula 1 data using the FastF1 library.
Tools to implement
- driver info
- compare drivers
"""
import json
import fastf1
import gradio as gr
import pandas as pd
from PIL import Image
from typing import Union
from fastf1.core import Session
# Local modules
from utils import parser_utils, track_utils
from utils.constants import (
AVAILABLE_SESSION_TYPES,
DRIVER_DETAILS,
CONSTRUCTOR_DETAILS
)
# Custom types
gp = Union[str, int]
session_type = Union[str, int, None]
### FastF1 tools ###
def get_session(year: int, round: gp, session_type: session_type) -> Session:
"""Retrieve a specific Formula 1 session.
Args:
year (int): The season year (e.g., 2024)
round (str | int): The race round number or name (e.g., 1 or 'Monaco')
session_type (str | int | None): Type of session (e.g., 'FP1', 'Q', 'R')
Returns:
Session: A FastF1 Session object for the specified parameters
Note:
If session_type is a string and not in AVAILABLE_SESSION_TYPES,
returns an error message string instead.
"""
# Check if session type is valid
if isinstance(session_type, str):
if session_type.lower() not in AVAILABLE_SESSION_TYPES:
return f"Session type {session_type} is not available. Supported session types: {AVAILABLE_SESSION_TYPES}"
if isinstance(round, str) and round.isnumeric():
round = int(round)
return fastf1.get_session(year, round, session_type)
def get_season_calendar(year: int) -> str:
"""Get the complete race calendar for a specific F1 season.
Args:
year (int): The season year to get the calendar for
Returns:
str: Formatted string containing the season calendar
"""
season_calendar = fastf1.get_event_schedule(year)
return parser_utils.parse_season_calendar(season_calendar)
def get_event_info(year: int, round: gp, format: str) -> str:
"""Retrieve information about a specific Formula 1 event.
Args:
year (int): The season year
round (str | int): The race round number or name
format (str): Output format ('human' for readable text, 'LLM' for structured data)
Returns:
str: Formatted event information based on the specified format
"""
if isinstance(round, str) and round.isnumeric():
round = int(round)
event = fastf1.get_session(year, round, "race").event # Event object is the same for all sessions, so hardcode "race"
if format == "human":
data_interval = f"{event['Session1DateUtc'].date()} - {event['Session5DateUtc'].date()}"
event_string = f"Round {event['RoundNumber']} : {event['EventName']} - {event['Location']}, {event['Country']} ({data_interval})"
return event_string
elif format == "LLM":
return parser_utils.parse_event_info(event)
def get_constructor_standings(year: int) -> str:
"""Retrieve constructor championship standings for a given year.
Args:
year (int): The season year
Returns:
str: Constructor championship standings
"""
pass
def get_driver_standings(year: int) -> str:
"""Retrieve driver championship standings for a given year.
Args:
year (int): The season year
Returns:
str: Driver championship standings
"""
pass
def driver_championship_standings(year: int, driver_name: str) -> str:
"""Get the championship standing for a specific driver in a given year.
Args:
year (int): The season year
driver_name (str): Full name of the driver (e.g., 'Lewis Hamilton')
Returns:
str: Formatted string with driver's position, points, and wins
"""
with open("assets/driver_abbreviations.json") as f:
driver_abbreviations = json.load(f)
driver_abbreviation = driver_abbreviations[driver_name]
ergast = fastf1.ergast.Ergast()
driver_standings = ergast.get_driver_standings(year).content[0]
driver_standing = driver_standings[["position", "points", "wins", "driverCode"]].reset_index(drop=True)
driver_standing = driver_standing[driver_standing["driverCode"] == driver_abbreviation]
suffix = "st" if driver_standing['position'].iloc[0] == 1 else "nd" if driver_standing['position'].iloc[0] == 2 else "rd" if driver_standing['position'].iloc[0] == 3 else "th"
standings_string = f"{driver_name} is {driver_standing['position'].iloc[0]}{suffix} with {driver_standing['points'].iloc[0]} points and {driver_standing['wins'].iloc[0]} wins"
return standings_string
def constructor_championship_standings(year: int, constructor_name: str) -> str:
"""Get the championship standing for a specific constructor in a given year.
Args:
year (int): The season year
constructor_name (str): Name of the constructor team (e.g., 'Mercedes')
Returns:
str: Formatted string with constructor's position, points, and wins
"""
team_mapping = {
"McLaren": "McLaren",
"Ferrari": "Ferrari",
"Red Bull Racing": "Red Bull",
"Mercedes": "Mercedes",
"Aston Martin": "Aston Martin",
"Alpine": "Alpine F1 Team",
"Haas": "Haas F1 Team",
"Racing Bulls": "RB F1 Team",
"Williams": "Williams",
"Kick Sauber": "Sauber"
}
ergast = fastf1.ergast.Ergast()
constructor_standings = ergast.get_constructor_standings(year).content[0]
constructor_standing = constructor_standings[["position", "points", "wins", "constructorName"]].reset_index(drop=True)
mapped_name = team_mapping[constructor_name]
constructor_standing = constructor_standing[constructor_standing["constructorName"] == mapped_name]
suffix = "st" if constructor_standing['position'].iloc[0] == 1 else "nd" if constructor_standing['position'].iloc[0] == 2 else "rd" if constructor_standing['position'].iloc[0] == 3 else "th"
standings_string = f"{constructor_name} are {constructor_standing['position'].iloc[0]}{suffix} with {constructor_standing['points'].iloc[0]} points and {constructor_standing['wins'].iloc[0]} wins"
return standings_string
def track_visualization(year: int, round: gp, visualization_type: str, driver_name: str) -> Image.Image:
"""Generate a visualization of the track with specified data.
Args:
year (int): The season year
round (str | int): The race round number or name
visualization_type (str): Type of visualization ('speed', 'corners', or 'gear')
driver_name (str): Name of the driver for driver-specific visualizations
Returns:
Image.Image: A PIL Image object containing the visualization
"""
if isinstance(round, str) and round.isnumeric():
round = int(round)
session = get_session(year, round, "race")
session.load()
if visualization_type == "speed":
return track_utils.create_track_speed_visualization(session, driver_name)
elif visualization_type == "corners":
return track_utils.create_track_corners_visualization(session)
elif visualization_type == "gear":
return track_utils.create_track_gear_visualization(session)
def get_session_results(year: int, round: gp, session_type: session_type) -> pd.DataFrame:
"""Retrieve and format the results of a specific session.
Args:
year (int): The season year
round (str | int): The race round number or name
session_type (str | int | None): Type of session (e.g., 'Q', 'R', 'Sprint')
Returns:
pd.DataFrame: DataFrame containing the session results
Raises:
ValueError: If the session type is invalid
"""
if isinstance(round, str) and round.isnumeric():
round = int(round)
try:
session = get_session(year, round, session_type)
session.load(telemetry=False)
# Create a proper copy of the results DataFrame
df = session.results[['DriverNumber', 'Abbreviation', 'FullName', 'Position',
'GridPosition', 'Points', 'Status', 'Q1', 'Q2', 'Q3']].copy()
except ValueError as e:
raise gr.Error(f"Session type {session_type} is not supported for the specified round. This Grand Prix most likely did not include a sprint race/quali.")
# Now we can safely modify the DataFrame
df["Name"] = df.apply(lambda row: f"{row['FullName']} ({row['Abbreviation']}{row['DriverNumber']})", axis=1)
df = df.drop(columns=["FullName", "Abbreviation", "DriverNumber"])
df = df.rename(columns={"Position": "Pos", "GridPosition": "Grid Pos"})
# Process results based on session type
if session_type in ["race", "sprint"]:
df = df[["Pos", "Name", "Points", "Grid Pos", "Status"]]
elif "qualifying" in session_type:
df[["Q1", "Q2", "Q3"]] = df[["Q1", "Q2", "Q3"]].apply(lambda x: x.dt.total_seconds().apply(lambda y: f"{int(y//60):02d}:{int(y%60):02d}.{int(y%1*1000):03d}" if pd.notna(y) else "-"))
df = df[["Pos", "Name", "Q1", "Q2", "Q3"]]
return df
def get_driver_info(driver_name: str) -> str:
"""Retrieve detailed information about a specific driver.
Args:
driver_name (str): Full name of the driver (e.g., 'Max Verstappen')
Returns:
str: Formatted string with driver's details including name, team, number,
nationality, and a brief summary
"""
driver = DRIVER_DETAILS[driver_name]
driver_info_string = f"{driver_name} ({driver['birth_date']}) {driver['nationality']}\n{driver['team']} #{driver['number']}\n\n{driver['summary']}"
return driver_info_string
def get_constructor_info(constructor_name: str) -> str:
"""Retrieve detailed information about a specific constructor.
Args:
constructor_name (str): Full name of the constructor (e.g., 'Red Bull Racing')
Returns:
str: Formatted string with constructor's details including name, team, number,
nationality, and a brief summary
"""
constructor = CONSTRUCTOR_DETAILS[constructor_name]
constructor_info_string = f"{constructor['team_name']} ({constructor_name})\n{constructor['base']}\nTeam principle: {constructor['team_principal']}\nDriver(s): {constructor['drivers'][0]} & {constructor['drivers'][1]}\nPower unit: {constructor['power_unit']}\nChassis: {constructor['chassis']}"
return constructor_info_string
if __name__ == "__main__":
session = get_session(2024, 1, "fp1")
session.load()