mean_absolute_percentage_error

mean_absolute_percentage_error#

mlproject.postprocess.utils.mean_absolute_percentage_error(y_true, y_pred, threshold=1e-5)[source]#

Compute mean absolute percentage error, masked

Masking is for when y_true is zero (causing a divide by zero error) or when y_true is very small (causing a massive skewing in the absolute percentage error).

Note: THIS WILL IGNORE ALL ENTRIES WHERE y_true’s MAGNITUDE IS less than the threshold, hence the MAPE score is not representative of all entries if the truth array contains entries with magnitude very close to 0.

Parameters:

  • y_true (np.ndarray) – A 1-D array of true values

  • y_pred (np.ndarray) – A 1-D array of predicted values

  • threshold (float) – Entries with magnitude below this value will be ignored in the output.

Returns:

Mean absolute percentage error, masked

Return type:

float

Contents