tabpfn-extensions
Conformal prediction with missing data
Compute the correction terms for missing data masks using conformal prediction.
View source
View source
View source
Was this page helpful?
Documentation Index
Fetch the complete documentation index at: /llms.txt
Use this file to discover all available pages before exploring further.
Compute the correction terms for missing data masks using conformal prediction.
pip install tabpfn-extensions
CPMDATabPFNRegressorCPMDATabPFNRegressor(
quantiles: list[float] | None = None,
val_size: float = 0.3,
seed: int | None = None,
tabpfn_estimator: TabPFNRegressor | None = None,
)
| Parameter | Type | Default | Description |
|---|---|---|---|
quantiles | list[float] | None | None | Three quantiles [lower, median, upper]. Must be symmetric (lower + upper == 1). |
val_ | float | 0.3 | Fraction of training data used for conformal calibration. |
seed | int | None | None | Random seed for the train/calibration split. |
tabpfn_ | Tab | None | A pre-configured TabPFNRegressor instance. If None, a default TabPFNRegressor() is used. Use this to customise parameters such as n_estimators, device, or random_state. Example::CPMDATabPFNRegressor( tabpfn_estimator=TabPFNRegressor(n_estimators=16, device=“cuda”) ) |
| Attribute | Type | Description |
|---|---|---|
model_ | Tab | Fitted TabPFN model. |
calibration_results_ | pd.Data | Correction terms for each observed missing-data mask. |
alpha_ | float | Miscoverage level derived from quantiles (lower * 2). |
feature_names_in_ | list | Column names seen during fit. |
CPMDATabPFNRegressor.fitCPMDATabPFNRegressor.fit(
x_train: ArrayLike,
y_train: ArrayLike,
) -> CPMDATabPFNRegressor
| Parameter | Type | Default | Description |
|---|---|---|---|
x_ | Array | Required | — |
y_ | Array | Required | — |
| Type | Description |
|---|---|
CPMDA | — |
CPMDATabPFNRegressor.predictCPMDATabPFNRegressor.predict(
x_new: ArrayLike,
) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray]
| Parameter | Type | Default | Description |
|---|---|---|---|
x_ | Array | Required | — |
| Type | Description |
|---|---|
tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray] | — |
Was this page helpful?