evaluate.eval_mae()

Mean absolute error using the forecast sample median as point estimate.

Usage

Source

evaluate.eval_mae(
    pred,
    truth,
)

Parameters

pred: Array

Forecast samples with the sample axis first.

truth: Array

Ground-truth values (matching pred without the sample axis).

Returns

float

The mean absolute error.