23 lines
809 B
Python
23 lines
809 B
Python
|
from .affiliation_bin.generics import convert_vector_to_events
|
|||
|
|
from .affiliation_bin.metrics import pr_from_events
|
|||
|
|
|
|||
|
|
|
|||
|
|
def evaluate(y_true: list, y_pred: list) -> float:
|
|||
|
|
"""
|
|||
|
|
F1PA评估方法，经过point adjust调整标签后再用F1评分
|
|||
|
|
:param y_true: 真实标签
|
|||
|
|
:param y_pred: 检测标签
|
|||
|
|
:return: affiliation的三个score
|
|||
|
|
"""
|
|||
|
|
true, pred = y_true.copy(), y_pred.copy()
|
|||
|
|
events_pred = convert_vector_to_events(pred)
|
|||
|
|
events_gt = convert_vector_to_events(true)
|
|||
|
|
Trange = (0, len(pred))
|
|||
|
|
|
|||
|
|
res = pr_from_events(events_pred, events_gt, Trange)
|
|||
|
|
aff_precision = res["precision"]
|
|||
|
|
aff_recall = res["recall"]
|
|||
|
|
if aff_recall == 0 or aff_precision == 0:
|
|||
|
|
return 0
|
|||
|
|
aff_f1 = 2 * aff_precision * aff_recall / (aff_precision + aff_recall)
|
|||
|
|
return aff_f1
|