pysteps.verification.detcatscores.det_cat_fct¶
-
pysteps.verification.detcatscores.
det_cat_fct
(pred, obs, thr, scores='', axis=None)¶ Calculate simple and skill scores for deterministic categorical (dichotomous) forecasts.
Parameters: - pred : array_like
Array of predictions. NaNs are ignored.
- obs : array_like
Array of verifying observations. NaNs are ignored.
- thr : float
The threshold that is applied to predictions and observations in order to define events vs no events (yes/no).
- scores : {string, list of strings}, optional
The name(s) of the scores. The default, scores=”“, will compute all available scores. The available score names are:
Name Description ACC accuracy (proportion correct) BIAS frequency bias CSI critical success index (threat score) F1 the harmonic mean of precision and sensitivity FA false alarm rate (prob. of false detection, fall-out, false positive rate) FAR false alarm ratio (false discovery rate) GSS Gilbert skill score (equitable threat score) HK Hanssen-Kuipers discriminant (Pierce skill score) HSS Heidke skill score MCC Matthews correlation coefficient POD probability of detection (hit rate, sensitivity, recall, true positive rate) SEDI symmetric extremal dependency index - axis : None or int or tuple of ints, optional
Axis or axes along which a score is integrated. The default, axis=None, will integrate all of the elements of the input arrays.
If axis is -1 (or any negative integer), the integration is not performed and scores are computed on all of the elements in the input arrays.
If axis is a tuple of ints, the integration is performed on all of the axes specified in the tuple.
Returns: - result : dict
Dictionary containing the verification results.