fit

BaseEstimator.fit(X: Union[ndarray, DataFrame], y: Union[ndarray, DataFrame], **kwargs) → None[source]

Fit the estimator to the data.

This method initializes the data loader, the model, and then starts the training process.

Parameters:

  • X (Union[np.ndarray, pd.core.frame.DataFrame]) – The input data.

  • y (Union[np.ndarray, pd.core.frame.DataFrame]) – The target data.

  • arguments (Optional keyword) –

  • eval (bool) – Print settings.

  • lr (float) – Learning rate for the optimizer.

  • betas (tuple) – Coefficients used for computing running averages of gradient and its square.

  • weight_decay (float) – Weight decay (L2 penalty).

  • eps (float) – Term added to the denominator to improve numerical stability.

  • amsgrad (bool) – Whether to use the AMSGrad variant of the Adam optimizer.

  • gamma (float) – Multiplicative factor of learning rate decay.

  • step_size (float) – Period of learning rate decay.

  • scheduler_patience (int) – The number of epochs to wait for improvement before stopping early.

  • min_delta (int) – Minimum change in the monitored quantity to qualify as an improvement.

  • scheduler (bool) – Whether to use a learning rate scheduler.

  • scheduler_mode (str) – The mode for the learning rate scheduler.

  • scheduler_factor (float) – The factor for reducing the learning rate.

  • scheduler_threshold (float) – The threshold for reducing the learning rate.

  • lrd (float) – Learning rate decay.

  • clip_norm (float) – Gradient clipping threshold.

  • scaler_type (str) – The type of scaler to use.

  • epochs (int) – The number of epochs to train for.

  • batch_size (int) – The size of the batches for training.

  • sequence_length (int) – The length of the sequence for training.

  • num_workers (int) – The number of worker threads to use for data loading.

  • validation_cadence (int) – The number of epochs between validation checks.

  • optim_args (float) – Additional arguments for the optimizer.

  • shuffle (bool) – Whether to shuffle the data before each epoch.

  • val_size (float) – The size of the validation set.

  • early_stopping (bool) – Whether to use early stopping.

  • pretrained (bool) – Whether to load a pre-trained model.

  • folder (str) – The folder to save the model to.

Returns:

None

predict

RegressorMixin.predict(X: Union[ndarray, DataFrame]) → ndarray[source]

Computes the prediction of the regressor on the given test data.

Parameters:

X (Union[np.ndarray, pd.core.frame.DataFrame]) – The test data.

Returns:

The predicted target.

Return type:

output (np.ndarray)

score

ClassifierMixin.score(X: Union[ndarray, DataFrame], y: Union[ndarray, DataFrame])[source]

Computes the score of the classifier on the given test data and labels.

Parameters:

  • X (Union[np.ndarray, pd.core.frame.DataFrame]) – The test data.

  • y (Union[np.ndarray, pd.core.frame.DataFrame]) – The true labels for the test data.

Returns:

The true labels. pred_labels (np.ndarray): The predicted labels.

Return type:

true_labels (np.ndarray)