[BUG] TeamId missing for BOR in 2025 São Paolo GP qualifying results

[mdsakalu]

Describe the issue:

For the 2025 São Paulo GP, Bortoleto missed the qualifying session due to crashing in the sprint race.

The ergast (jolpica) data does not have an entry for Bortoleto, while the F1 API data does, which results in the TeamId being nan.

{'Abbreviation': 'BOR',
 'BroadcastName': 'G BORTOLETO',
 'ClassifiedPosition': '',
 'CountryCode': '',
 'DriverId': 'nan',
 'DriverNumber': '5',
 'FirstName': 'Gabriel',
 'FullName': 'Gabriel Bortoleto',
 'GridPosition': nan,
 'HeadshotUrl': 'https://media.formula1.com/d_driver_fallback_image.png/content/dam/fom-website/drivers/G/GABBOR01_Gabriel_Bortoleto/gabbor01.png.transform/1col/image.png',
 'Laps': nan,
 'LastName': 'Bortoleto',
 'Points': nan,
 'Position': nan,
 'Q1': NaT,
 'Q2': NaT,
 'Q3': NaT,
 'Status': '',
 'TeamColor': '01C00E',
 'TeamId': 'nan',
 'TeamName': 'Kick Sauber',
 'Time': NaT}

Looking at the code, it seems like it's set up for the opposite scenario where the F1 API data would be missing the row.

Since TeamName is in the F1 API data, this could be fixed by taking the TeamId from the teammate's row.

if 'TeamId' in self._results and 'TeamName' in self._results:
    # Build a mapping from drivers who have both TeamName (F1 API) and TeamId (Ergast) already populated
    team_id_map = {}
    for idx in self._results.index:
        team_name = self._results.loc[idx, 'TeamName']
        team_id = self._results.loc[idx, 'TeamId']
        # Check for valid (non-NaN, non-'nan' string) values
        if (pd.notna(team_name) and pd.notna(team_id) and str(team_id).lower() != 'nan'):
            team_id_map[team_name] = team_id

        # Apply the mapping to fill NaN/missing TeamId values
         mask = ((self._results['TeamId'].isna()
             | (self._results['TeamId'].astype(str).str.lower() == 'nan'))
             & self._results['TeamName'].notna())
         for idx in self._results[mask].index:
             team_name = self._results.loc[idx, 'TeamName']
             if team_name in team_id_map:
                 self._results.loc[idx, 'TeamId'] = team_id_map[team_name]

Reproduce the code example:

import fastf1
from pprint import pprint

# Reproduce: Bortoleto TeamId is NaN for 2025 São Paulo GP qualifying
# Reason: He's registered but didn't participate in the session

event = fastf1.get_event(2025, 21)
q = event.get_qualifying()
q.load()

# Check Bortoleto's data
bort = q.results[q.results['Abbreviation'] == 'BOR'].iloc[0]
pprint(bort.to_dict())

Error message:

[theOehrly]

This is an unfortunate problem, yes there is a reasonable data source here, but this is very much an edge case. There are likely many potential edge cases like this. Adding tailor made fixes like this will take time and will be difficult to maintain.

I'll keep this open as a bug, but this needs to be considered in the context of broader improvements to the data loading and data augmentation situation.

[CodersAcademy006]

I’ve opened a PR that addresses this issue by adding explicit warnings for known, recoverable data inconsistencies and refactoring the related utilities to keep the warning infrastructure clean and maintainable; the change is fully backward-compatible and improves transparency without altering behavior. If any further changes, please review and provide me the feedback