Filtering

Pyrox filters at read time, so you load only the rows you need.

Gender and division

These filters are applied server-side for efficiency.

race = client.get_race(
    season=7,
    location="london",
    gender="male",      # "male" | "female" | "mixed"
    division="open",
)

Supported divisions are open, pro, doubles, pro_doubles, relay, adaptive, elite and elite_doubles.

Notes: - Values are case-sensitive in the underlying parquet filter; prefer lowercase. - If the filtered dataset has no rows, Pyrox raises RaceNotFound.

Elite results

Use division="elite" for Elite singles or division="elite_doubles" for pairs. They are separate from pro and pro_doubles; not every race offers them.

elite = client.get_race(
    season=8,
    location="hamburg",
    year=2025,
    division="elite",  # "elite_doubles" for pairs
    use_cache=False,
)

Season 8 includes 461 Elite singles and doubles results across Hamburg (2025), Melbourne (2025), Phoenix (2026), EMEA London Olympia (2026), APAC Championship Brisbane (2026), Warsaw (2026), Stockholm (2026) and Washington DC (2026). Use client.list_races(season=8) to find the location slugs and years.

Published fractional seconds are preserved when times are converted to numeric minutes. Missing source splits remain missing. These filters work with pyrox-client 0.2.7; no package upgrade is needed for the new data. use_cache=False downloads fresh results instead of reusing a cached race.

Total time windows

total_time is expressed in minutes. You can pass a single value or an open interval.

# Under 60 minutes
sub_60 = client.get_race(season=7, location="london", total_time=60)

# Open interval: 50 < total_time < 60
mid_pack = client.get_race(season=7, location="london", total_time=(50, 60))

# Only lower bound
slow = client.get_race(season=7, location="london", total_time=(70, None))

Notes: - Bounds are strict (> and <), not inclusive. - Filtering happens after time columns are converted to minutes.