tabensemb.utils.utils.auto_metric_sklearn#

tabensemb.utils.utils.auto_metric_sklearn(y_true: ndarray, y_pred: ndarray, metric: str, task: str) → float[source]#

Calculate metrics using sklearn APIs. It automatically deals with different requirements of input shapes for different metrics.

Parameters:

y_true: An array of ground truth values. For classification, it should be the class of each sample. It can be 1d or 2d (the second dimension is 1) for classification tasks.
y_pred: An array of predictions. For classification, it should be the probabilities of classes. It can be 1d or 2d (the second dimension is 1) for binary classification tasks.
metric: Use tabensemb.utils.utils.REGRESSION_METRICS, tabensemb.utils.utils.BINARY_METRICS, and tabensemb.utils.utils.MULTICLASS_METRICS to check all available metrics for regression, binary, and multiclass tasks respectively.
task: “regression”, “multiclass”, or “binary”.

Returns: