Fitting and prediction

Looking for usage documentation? Check out Classification.

`TabPFNClassifier.fit`

Fit the model.

TabPFNClassifier.fit(
    X: XType,
    y: YType,
) -> Self

Type aliases

XType = Any
YType = Any

Parameters

Parameter	Type	Default	Description
`X`	`XType`	Required	The input data.
`y`	`YType`	Required	The target variable.

Returns

Type	Description
`Self`	self

`TabPFNClassifier.predict`

View source

Predict the class labels for the provided input samples.

TabPFNClassifier.predict(
    X: XType,
) -> np.ndarray

Type aliases

XType = Any

Parameters

Parameter	Type	Default	Description
`X`	`XType`	Required	The input data for prediction.

Returns

Type	Description
`np.ndarray`	The predicted class labels as a NumPy array.

`TabPFNClassifier.predict_logits`

View source

Predict the raw logits for the provided input samples. Logits represent the unnormalized log-probabilities of the classes before the softmax activation function is applied.

TabPFNClassifier.predict_logits(
    X: XType,
) -> np.ndarray

Type aliases

XType = Any

Parameters

Parameter	Type	Default	Description
`X`	`XType`	Required	The input data for prediction.

Returns

Type	Description
`np.ndarray`	The predicted logits as a NumPy array. Shape (n_samples, n_classes).

`TabPFNClassifier.predict_raw_logits`

View source

Predict the raw logits for the provided input samples. Logits represent the unnormalized log-probabilities of the classes before the softmax activation function is applied. In contrast to the predict_logits method, this method returns the raw logits for each estimator, without averaging estimators or temperature scaling.

TabPFNClassifier.predict_raw_logits(
    X: XType,
) -> np.ndarray

Type aliases

XType = Any

Parameters

Parameter	Type	Default	Description
`X`	`XType`	Required	The input data for prediction.

Returns

Type	Description
`np.ndarray`	An array of predicted logits for each estimator, Shape (n_estimators, n_samples, n_classes).

`TabPFNClassifier.predict_proba`

View source

Predict the probabilities of the classes for the provided input samples. This is a wrapper around the _predict_proba method.

TabPFNClassifier.predict_proba(
    X: XType,
) -> np.ndarray

Type aliases

XType = Any

Parameters

Parameter	Type	Default	Description
`X`	`XType`	Required	The input data for prediction.

Returns

Type	Description
`np.ndarray`	The predicted probabilities of the classes as a NumPy array. Shape (n_samples, n_classes).

Scikit-learn scoring

See inherited score.

Reference

tabpfn

tabpfn-client

tabpfn-extensions

REST API

Fitting and prediction

`TabPFNClassifier.fit`

`TabPFNClassifier.predict`

`TabPFNClassifier.predict_logits`

`TabPFNClassifier.predict_raw_logits`

`TabPFNClassifier.predict_proba`

Scikit-learn scoring

​TabPFNClassifier.fit

​TabPFNClassifier.predict

​TabPFNClassifier.predict_logits

​TabPFNClassifier.predict_raw_logits

​TabPFNClassifier.predict_proba

​Scikit-learn scoring

`TabPFNClassifier.fit`

`TabPFNClassifier.predict`

`TabPFNClassifier.predict_logits`

`TabPFNClassifier.predict_raw_logits`

`TabPFNClassifier.predict_proba`

Scikit-learn scoring