Datasets
v0.1.7Bundled datasets ready for training, with the same loader signature as StatsLibX and ViewX — including polars backend and (X, y) array output.
Bundled Datasets
| Dataset | Description | Rows | Cols |
|---|---|---|---|
| iris | Fisher's Iris flower measurements (3 species) | 150 | 5 |
| penguins | Palmer penguins morphometrics (3 species) | 344 | 7 |
| titanic | Titanic passenger manifest | 418 | 12 |
| sp500_companies | S&P 500 companies with financial metrics | 503 | 21 |
| course_completion | Student course completion records | 100,000 | 40 |
API Reference
load_dataset
load_dataset(name, backend='pandas', return_X_y=None) -> pd.DataFrame | tuple[np.ndarray, np.ndarray]
pd.DataFrame | pl.DataFrame | tuple[np.ndarray, np.ndarray]
list_datasets
list_datasets() -> list[str]
list[str]
Ecosystem Compatibility
The same load_dataset call works across the ecosystem, and polars output from StatsLibX feeds SynaptIX models directly:
cross.py
python
from statslibx import load_dataset as slx_load
import synaptix as sx
df_pl = slx_load("iris.csv", backend="polars") # statslibx polars backend
model = sx.supervised.RandomForestClassifier()
model.fit(df_pl.drop("species"), df_pl["species"]) # converted internally