graphtoolbox.training.metrics

Functions

BIAS(preds, targets)

Bias between predictions and targets.

MAE(preds, targets)

Mean Absolute Error (MAE) between predictions and targets.

MAPE(preds, targets)

Mean Absolute Percentage Error (MAPE) between predictions and targets.

NMAE(preds, targets)

Normalized Mean Absolute Error (NMAE) between predictions and targets.

RMSE(preds, targets)

Root Mean Square Error (RMSE) between predictions and targets.
graphtoolbox.training.metrics.MAE(preds: Tensor | ndarray, targets: Tensor | ndarray) → Tensor | float[source][source]

Mean Absolute Error (MAE) between predictions and targets.

Parameters:

  • preds (Union[torch.Tensor, np.ndarray]) – Predicted values.

  • targets (Union[torch.Tensor, np.ndarray]) – True values.

Returns:

The mean absolute error.

Return type:

Union[torch.Tensor, float]

graphtoolbox.training.metrics.NMAE(preds: Tensor | ndarray, targets: Tensor | ndarray) → Tensor | float[source][source]

Normalized Mean Absolute Error (NMAE) between predictions and targets.

Parameters:

  • preds (Union[torch.Tensor, np.ndarray]) – Predicted values.

  • targets (Union[torch.Tensor, np.ndarray]) – True values.

Returns:

The normalized mean absolute error.

Return type:

Union[torch.Tensor, float]

graphtoolbox.training.metrics.MAPE(preds: Tensor | ndarray, targets: Tensor | ndarray) → Tensor | float[source][source]

Mean Absolute Percentage Error (MAPE) between predictions and targets.

Parameters:

  • preds (Union[torch.Tensor, np.ndarray]) – Predicted values.

  • targets (Union[torch.Tensor, np.ndarray]) – True values.

Returns:

The mean absolute percentage error.

Return type:

Union[torch.Tensor, float]

graphtoolbox.training.metrics.RMSE(preds: Tensor | ndarray, targets: Tensor | ndarray) → Tensor | float[source][source]

Root Mean Square Error (RMSE) between predictions and targets.

Parameters:

  • preds (Union[torch.Tensor, np.ndarray]) – Predicted values.

  • targets (Union[torch.Tensor, np.ndarray]) – True values.

Returns:

The root mean square error.

Return type:

Union[torch.Tensor, float]

graphtoolbox.training.metrics.BIAS(preds: Tensor | ndarray, targets: Tensor | ndarray) → Tensor | float[source][source]

Bias between predictions and targets.

Parameters:

  • preds (Union[torch.Tensor, np.ndarray]) – Predicted values.

  • targets (Union[torch.Tensor, np.ndarray]) – True values.

Returns:

The bias (mean error).

Return type:

Union[torch.Tensor, float]