yeyupiaoling / speechemotionrecognition-paddlepaddle Goto Github PK

for x_val, y_val in zip(x.flatten(), y.flatten()):
c = cm[y_val][x_val] / (np.sum(cm[:, x_val]) + 1e-6)
# 忽略值太小的
if c < 1e-4:continue
plt.text(x_val, y_val, "%0.2f" % (c,), color='red', fontsize=15, va='center', ha='center')
报错内容
Traceback (most recent call last):
File "D:\SpeechEmotionRecognition-Pytorch-master\SpeechEmotionRecognition-Pytorch-master\eval.py", line 71, in
evaluate()
File "D:\SpeechEmotionRecognition-Pytorch-master\SpeechEmotionRecognition-Pytorch-master\eval.py", line 66, in evaluate
plot_confusion_matrix(cm=cm, save_path='output/log/混淆矩阵_eval.png', class_labels=class_labels, show=False)
File "D:\SpeechEmotionRecognition-Pytorch-master\SpeechEmotionRecognition-Pytorch-master\utils\utility.py", line 31, in plot_confusion_matrix
c = cm[y_val][x_val] / (np.sum(cm[:, x_val]) + 1e-6)
IndexError: index 5 is out of bounds for axis 0 with size 5

index = magnitudes[:, 1].argmax()
IndexError: index 1 is out of bounds for axis 1 with size 1

我把音频文件名改成了300_12_32.wav的格式之后，就报了错，想知道原因是什么？求解