Metrics and scoring#

Functions for evaluating the quality of a model’s predictions.

equisolve.utils.metrics.rmse(y_true: TensorMap, y_pred: TensorMap, parameter_key: str) → List[float][source]#

Mean squared error regression loss.

Parameters:

  • y_true – Ground truth (correct) target values.

  • y_pred – Estimated target values.

  • parameter_key – Parameter to perform the rmse for. Examples are "values", "positions" or "cell".

Returns loss:

A non-negative floating point value (the best value is 0.0), or a tuple of floating point values, one for each block in y_pred.