cleanskate 0.1.0

Pandas-friendly access to figure skating scores

⛸️ cleanskate

cleanskate is a Python package for loading figure skating scores as pandas data frames. The package covers overall results, judge information, program component scores, and per-element scores for international events from the 2018-2019 season onward.

The current dataset covers major international events, including:

• Olympics
• Worlds
• Junior Worlds
• Europeans
• Four Continents
• Grand Prix
• Grand Prix Final
• Junior Grand Prix
• much of the Challenger Series

The dataset may be expanded in the future. Feel free to open an issue if you notice anything missing.

Installation

pip install cleanskate

For local development:

pip install -e .

Quick Start

from cleanskate import Dataset

ds = Dataset(version="latest")

events = ds.load_events()
results = ds.load_results()
elements = ds.load_elements()

Tables are downloaded automatically on first use and cached locally. You do not need to call a separate download command.

Common Filters

All loader filters accept either a single value or a list of values. Lists use "one of these values" semantics.

from cleanskate import Dataset

ds = Dataset()

worlds_events = ds.load_events(event_series="Worlds")
season_results = ds.load_results(season="2025-2026")
women_segments = ds.load_segments(discipline="Women")
triple_axels = ds.load_elements(attempt_code="3A")

senior_jump_attempts = ds.load_elements(
    event_level="Senior",
    element_family="Jump",
)

non_clean_jumps = ds.load_elements(
    attempt_code=["3A", "4T", "4S"],
    clean_element=False,
)

Recommended public filters:

• season
• event_series
• event_level
• event_label
• segment_label
• discipline
• element_family
• attempt_code
• clean_element

Lower-level IDs like event_id, segment_id, and result_id are also available for power users.

Local Datasets

You can point Dataset at a local directory instead of the hosted snapshot:

from cleanskate import Dataset

ds = Dataset(base_dir="/path/to/local/dataset")
segments = ds.load_segments()

prefetch() is available if you want to warm the cache explicitly:

ds.prefetch()

Example Notebooks

The repository currently includes:

• Quick start notebook
• Judge nationality patterns notebook

Documentation

Additional docs:

• Getting started
• API reference
• Data model
• Dataset operations
• Dataset changelog
• Provenance and citation
• Analysis notes
• Release process
• Release readiness

Related Work

• BuzzFeed data parsers
• BuzzFeed judge bias analysis