pysteps.verification.probscores.ROC_curve

pysteps.verification.probscores.ROC_curve(P_f, X_o, X_min, n_prob_thrs=10, compute_area=False)

Compute the ROC curve and its area from the given ROC object.

Parameters:

P_f : array_like

Forecasted probabilities for exceeding the threshold specified in the ROC object. Non-finite values are ignored.

X_o : array_like

Observed values. Non-finite values are ignored.

X_min : float

Precipitation intensity threshold for yes/no prediction.

n_prob_thrs : int

The number of probability thresholds to use. The interval [0,1] is divided into n_prob_thrs evenly spaced values.

compute_area : bool

If True, compute the area under the ROC curve (between 0.5 and 1).

Returns:

out : tuple

A two-element tuple containing the probability of detection (POD) and probability of false detection (POFD) for the probability thresholds specified in the ROC curve object. If compute_area is True, return the area under the ROC curve as the third element of the tuple.