Source code for huracanpy.assess.scores

"""Functions for score computation"""

import numpy as np



[docs]
def POD(matches, ref, ref_name="ib"):
    """

    Parameters
    ----------
    matches
    ref
    ref_name

    Returns
    -------

    """
    N_detected = matches["id_" + ref_name].nunique()
    N_total = len(np.unique(ref.track_id.values))
    return N_detected / N_total




[docs]
def FAR(matches, detected, detected_name="UZ"):
    """

    Parameters
    ----------
    matches
    detected
    detected_name

    Returns
    -------

    """
    N_matched = matches["id_" + detected_name].nunique()
    N_total = len(np.unique(detected.track_id.values))
    return 1 - (N_matched / N_total)