Python 为什么我要找一个索引器来获得最佳精度阈值？

我在另一台机器上运行了以下代码，但在另一台机器上运行时遇到以下错误：

class_names = ['Fish', 'Flower', 'Sugar', 'Gravel']

def get_threshold_for_recall(y_true, y_pred, class_i, recall_threshold=0.94, precision_threshold=0.90, plot=False):

    precision, recall, thresholds = precision_recall_curve(y_true[:, class_i], y_pred[:, class_i])
    i = len(thresholds) - 1
    best_recall_threshold = None
    while best_recall_threshold is None:
        next_threshold = thresholds[i]
        next_recall = recall[i]
        if next_recall >= recall_threshold:
            best_recall_threshold = next_threshold
        i -= 1

    # consice, even though unnecessary passing through all the values
    best_precision_threshold = [thres for prec, thres in zip(precision, thresholds) if prec >= precision_threshold][0]

    if plot:
        plt.figure(figsize=(10, 7))
        plt.step(recall, precision, color='r', alpha=0.3, where='post')
        plt.fill_between(recall, precision, alpha=0.3, color='r')
        plt.axhline(y=precision[i + 1])
        recall_for_prec_thres = [rec for rec, thres in zip(recall, thresholds) 
                                 if thres == best_precision_threshold][0]
        plt.axvline(x=recall_for_prec_thres, color='g')
        plt.xlabel('Recall')
        plt.ylabel('Precision')
        plt.ylim([0.0, 1.05])
        plt.xlim([0.0, 1.0])
        plt.legend(['PR curve', 
                    f'Precision {precision[i + 1]: .2f} corresponding to selected recall threshold',
                    f'Recall {recall_for_prec_thres: .2f} corresponding to selected precision threshold'])
        plt.title(f'Precision-Recall curve for Class {class_names[class_i]}')
    return best_recall_threshold, best_precision_threshold


y_pred = model.predict_generator(data_generator_val, workers=num_cores)
y_true = data_generator_val.get_labels()
recall_thresholds = dict()
precision_thresholds = dict()

for i, class_name in tqdm(enumerate(class_names)):
    recall_thresholds[class_name], precision_thresholds[class_name] = get_threshold_for_recall(y_true, y_pred, i, plot=True)

我希望这四个类有四条精度召回曲线，但是，我得到了以下错误消息：

---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
<ipython-input-79-422044a4f5da> in <module>
     37 precision_thresholds = dict()
     38 for i, class_name in tqdm(enumerate(class_names)):
---> 39     recall_thresholds[class_name], precision_thresholds[class_name] = get_threshold_for_recall(y_true, y_pred, i, plot=True)

<ipython-input-79-422044a4f5da> in get_threshold_for_recall(y_true, y_pred, class_i, recall_threshold, precision_threshold, plot)
     12 
     13     # consice, even though unnecessary passing through all the values
---> 14     best_precision_threshold = [thres for prec, thres in zip(precision, thresholds) if prec > precision_threshold][0]
     15 
     16     if plot:

IndexError: list index out of range

---------------------------------------------------------------------------
索引器回溯（最后一次最近调用）
在里面
37精度_阈值=dict（）
38对于i，TQM中的类名称（枚举（类名称））：
--->39调用阈值[类名称]，精度阈值[类名称]=获取调用阈值（y\u真，y\u pred，i，plot=真）
用于调用的获取阈值（y\u真、y\u pred、类i、调用阈值、精度阈值、绘图）
12
13#一致性，即使不需要通过所有值
--->14最佳精度阈值=[prec的阈值，如果prec>精度阈值，则zip中的阈值（精度，阈值）[0]
15
16如果绘图：
索引器：列表索引超出范围

解决了我在上面的代码片段中遇到的问题。请确保您有正确的版本：tensorflow==1.14.0和keras=2.3.0

显然，

[thres for prec，thres in zip（精度，阈值），如果prec>precision\u threshold]

正在创建一个空列表。