FastF1 Documentation

FastF1 provides programmatic access to Formula 1 data, including lap timing, car telemetry, position, tyre data, weather data, event schedules, and session results.

Key features include:

• Pandas Integration: All data is provided as extended Pandas DataFrames, including custom functions specifically designed for F1 data manipulation.
• Ergast Compatibility: Full support for the Ergast-compatible jolpica-f1 API for accessing current and historical data.
• Visualization: Built-in integration with Matplotlib for data visualization.
• Caching: Implements caching for all API requests to improve performance. Since sessions can involve large data chunks (50-100MB), caching is enabled by default using an OS-specific location. You can configure a custom cache location using fastf1.req.Cache.

FastF1 uses a caching and rate limiting system for all HTTP requests.

• Caching: Enabled by default. It speeds up programs by serving requests from local storage. Disabling it is highly discouraged as it significantly slows down execution.
• Rate Limits: Applied to all HTTP requests. However, requests served from the cache do not count towards rate limits. Enabling the cache effectively increases your available rate limit.
• Rate Limit Enforcement:
  • Soft limit: FastF1 will throttle the request rate by introducing small delays.
  • Hard limit: FastF1 will raise a fastf1.exceptions.RateLimitExceededError.

The FastF1 API is categorized into the following functional areas to help you find the right tools for your task:

• Functions: Data loading (loading_data) and data plotting (plotting_data).
• Objects: Core data entities including events, session, timing_data, telemetry, results, and circuit_info.
• External APIs: Integrations with jolpica and ergast.
• Advanced Usage: Specialized modules for accounts_auth, cache_and_rate_limits, exceptions, logging, livetiming, utils, and deprecated_legacy features.

All classes and functions exposed in the fastf1.* namespace are public and available for use. The library is organized into several functional subpackages:

fastf1.core
fastf1.ergast
fastf1.events
fastf1.legacy
fastf1.livetiming
fastf1.plotting
fastf1.utils

In version 2.1.8, temporary support was added for sprint qualifying sessions. To retrieve data for a sprint qualifying session, use the fastf1.get_session function and pass event='SQ' as an argument.

Note for Sprint Weekends: On weekends featuring sprint qualifying, the session order changes:

• FP2 occurs on Saturday.
• FP3 does not exist.

Instead of using round numbers, you can load events or sessions using descriptive names (e.g., 'French Grand Prix', 'Spain') or locations (e.g., 'Silverstone'). FastF1 performs a best-match search, so ensure names are precise to avoid incorrect matches (e.g., 'Emilian' might match 'Belgian Grand Prix' instead of 'Emilia Romagna').

import fastf1

# Load event by name
event = fastf1.get_event(2021, 'French Grand Prix')

# Load session by location and session type
session = fastf1.get_session(2021, 'Silverstone', 'Q')

The method fastf1.core.Session.load_laps is deprecated. You should use fastf1.core.Session.load instead, which provides more flexibility for selecting data to load.

Important: The new load method does not return a fastf1.core.Laps object. Instead, you should access laps via the fastf1.core.Session.laps attribute.

The IsAccurate boolean flag indicates if a lap's start and end times are correctly synced with other laps. A lap marked as IsAccurate == True has passed these criteria:

• It is not an inlap or outlap.
• It was set under green or yellow flag conditions.
• It is not the first lap immediately following a safety car (SC/VSC) period.
• It contains a valid LapTime and all SectorTime values.
• The sum of the sector times matches the LapTime.

Note: Laps marked as inaccurate should be handled with caution as they may contain errors that are difficult to detect.

As of version 3.8.0, the minimum supported Python version is 3.10. Support for Python 3.9 has been dropped.

Additionally, the following minimum dependency versions are required:

• matplotlib>=3.8.0
• numpy>=1.26.0
• pandas>=2.1.1
• requests>=2.30.0
• scipy>=1.11.0
• pydantic (new dependency)