Introduction

Learn about tif1 - a fast Formula 1 data library with fastf1-compatible API

What is tif1?

tif1 is a fast, canonical Formula 1 data library that provides access to comprehensive F1 timing data from 2018 to the present. It fetches data from TracingInsights’ canonical dataset and offers a familiar API surface compatible with fastf1.

Key Insight

tif1 is built from the ground up for speed. While data becomes available approximately 30 minutes after a session ends (slightly longer than fastf1’s 20-25 minutes), this delay reflects the additional data enrichment and processing that makes tif1’s dataset more comprehensive.

Why tif1?

tif1 was created to address specific pain points in F1 data analysis:

⚡ Performance First

Direct CDN access with SQLite caching. Fetch specific lap telemetry in seconds without loading entire sessions.

🎯 Lazy Loading

No session.load() needed. Data is fetched on-demand, only when you access it.

🚀 Parallel Fetching

Async HTTP requests fetch multiple telemetry streams simultaneously, achieving 4-5x speedup.

💪 Reliable

Built-in retry logic with circuit breaker pattern and automatic CDN fallback.

Key Features

Speed & Performance

• Direct CDN Access: Data served via jsDelivr CDN with minimal latency
• SQLite Cache: Intelligent caching with JSON storage for instant repeated access
• Async Parallelism: Fetch telemetry for 19 drivers in ~0.4s (vs ~1.3s sequential)
• Lazy Evaluation: Only fetch what you need, when you need it

Data Completeness

Access comprehensive Formula 1 data including:

• Lap times with sector splits (S1, S2, S3)
• Tire compounds and stint information
• Telemetry: Speed, Throttle, Brake, RPM, Gear, DRS
• Position data (X, Y, Z coordinates)
• Acceleration data (X, Y, Z axes)
• Distance and relative distance
• Weather conditions
• Race control messages

Developer Experience

1. Type Safety

   Comprehensive type hints throughout for excellent IDE autocomplete and type checking

2. Error Handling

   Rich exception hierarchy (DataNotFoundError, DriverNotFoundError, NetworkError, etc.)

3. Jupyter Ready

   Built-in rich HTML display for notebooks with formatted DataFrames

4. Configurable

   .tif1rc configuration file support for project-specific settings

tif1 vs fastf1

tif1 is designed as a performance-focused alternative to fastf1 with a compatible API surface.

When to Use tif1

✅ Choose tif1 when you need:

• Fast access to specific lap telemetry without loading entire sessions
• Parallel data fetching for analysis across multiple drivers
• Minimal memory footprint with lazy loading
• Modern async/await patterns in your data pipelines
• Optional Polars backend for 2x faster DataFrame operations

When to Use fastf1

✅ Choose fastf1 when you need:

• Live timing data during sessions
• Ergast API integration for historical statistics
• Data available 5-10 minutes earlier after session ends
• Proven stability with years of production use

API Compatibility

tif1 implements fastf1’s core API:

import tif1

# Same familiar API
session = tif1.get_session(2025, "Monaco Grand Prix", "Qualifying")
laps = session.laps
driver = session.get_driver("VER")
fastest = driver.get_fastest_lap()
telemetry = fastest.telemetry

Most fastf1 code works with tif1 by simply changing the import statement. See the fastf1 Compliance Matrix for detailed compatibility.

What’s Different?

Out of Scope:

• fastf1.ergast - Historical statistics API
• fastf1.livetiming - Live session data streaming

Data Source:

• tif1 uses TracingInsights canonical data (GitHub via jsDelivr CDN)
• fastf1 uses official FIA API and timing data

Performance Examples

Here’s what tif1’s performance optimizations enable:

Sequential (Slow)

# Traditional sequential fetching
for driver_code in ["VER", "HAM", "LEC"]:
    driver = session.get_driver(driver_code)
    fastest = driver.get_fastest_lap()
    tel = fastest.telemetry  # ~0.4s each = 1.2s total

Data Availability

Data becomes available approximately 30 minutes after a session ends. This is 5-10 minutes longer than fastf1 due to additional enrichment processing.

Supported Years: 2018 - Present

Session Types:

• Practice 1, Practice 2, Practice 3
• Qualifying
• Sprint, Sprint Qualifying, Sprint Shootout
• Race

Architecture Highlights

tif1 is optimized for speed at every layer:

• HTTP Layer: niquests with connection pooling and keep-alive
• Async Engine: nest-asyncio for concurrent fetching in any context
• JSON Parsing: orjson for fast C-based parsing
• Caching: SQLite with in-memory buffer for sub-millisecond lookups
• Circuit Breaker: Automatic failure detection with exponential backoff
• CDN Strategy: Primary jsDelivr with automatic fallback

Next Steps

Installation

Install tif1 and optional dependencies

Quick Start

Get started with your first tif1 script