CorneringAnalyzer
Corner-by-corner analysis with entry/apex/exit speeds, brake points, and time loss
Configuration
- Slice Type:
lap - Metadata Only: False
- Required Channels:
speed,distance,time
When to Use
- User asks about specific corners or turns
- User wants to know where they're losing time
- User asks about corner entry, apex, or exit performance
- User asks about consistency through corners
Options
| Option | Type | Default | Description |
|---|---|---|---|
granularity | Literal[coarse, normal, detailed] | "normal" | Level of detail in analysis output |
comparison_mode | Literal[absolute, relative, percentage] | "absolute" | How to compare metrics across subjects |
landmark_kind | str | "turn" | Type of landmark to analyze (turn, turn_full_track, etc.) |
corner_names | list[str] | None | "—" | Specific corner names to analyze (None = all corners) |
brake_threshold | float | 0.1 | Minimum brake pressure to detect brake point Constraints: ≥ 0.0, ≤ 1.0 |
throttle_threshold | float | 0.2 | Minimum throttle to detect throttle application Constraints: ≥ 0.0, ≤ 1.0 |
lap_selection | str | "all" | Lap selection mode: 'all' for all laps, 'best' for single best lap, 'n_best' for top N laps Choices: all, best, n_best |
n_best_laps | int | None | 5 | Number of best laps to analyze when using 'n_best' lap selection mode |
min_valid_speed_kmh | float | 10.0 | Minimum valid speed threshold (km/h) - exclude laps with speeds below this Constraints: ≥ 0.0 |
sort_order | str | "time_loss" | How to sort corners: 'time_loss' (biggest problems first), 'track_order' (sequential) Choices: time_loss, track_order |
Examples
Example 1
User Query: Where am I losing time at Spa?
Call:
analyze(analyzers='cornering', track='Spa', lap_selection='best', sort_order='time_loss')
Explanation: Analyzes best lap, ranks corners by time loss to show biggest opportunities
Example 2
User Query: How consistent am I across all my laps in that session?
Call:
analyze(analyzers='cornering', event='01H8SP4HK4', lap_selection='all')
Explanation: Analyzes all laps to show consistency metrics (std dev, CV%) per corner
Example 3
User Query: Show me my performance through Big Bend
Call:
analyze(analyzers='cornering', track='Lime Rock', corner_names=['big_bend'], sort_order='track_order')
Explanation: Filters analysis to specific corner with sequential ordering