mean_absolute_error

rojak.turbulence.metrics.mean_absolute_error(truth: Array, prediction: Array) → float

Mean Absolute Error (MAE)

\[MAE(y, \hat{y}) = \frac{1}{n_{samples}} \sum_{i=0}^{n_{samples} - 1} | y_i - \hat{y_i} |\]

where \(n_{samples}\) is the number of samples, \(y_i\) is the truth value and \(\hat{y_i}\) is the corresponding predicted value.

Parameters:

• truth (Array)

• prediction (Array)

Return type:

float

Returns:

Examples

This example is modified from the scikit-learn’s user guide on MAE

>>> y_true = da.asarray([3, -0.5, 2, 7])
>>> y_pred = da.asarray([2.5, 0.0, 2, 8])
>>> float(mean_absolute_error(y_true, y_pred).compute())
0.5