python - 'soft' 错误的异常处理

Question

我在跑 means2来自 scipy，虽然我收到一条错误消息:

/usr/lib/python2.7/dist-packages/scipy/cluster/vq.py:600: UserWarning: One of the clusters is empty. Re-run kmean with a different initialization. warnings.warn("One of the clusters is empty. "

在循环中运行以下代码时:

training_2= numpy.random.random_integers(0, 400, size=[50,1]).astype(numpy.float32)
cent, clus= kmeans2(training_2, 3, minit='points')

循环不会终止。我也注意到 numpy 和 OpenCV 的这种行为。

通常的 try except block 不会捕获这些(因为缺少更好的术语)“软”错误；还有另一种方法来处理这些错误吗？

如果这不能直接实现，那么有没有办法让 python 脚本在运行时读取它自己的输出？

Answer 1

这不是一个异常(exception)，它是一个warning :

Warning messages are typically issued in situations where it is useful to alert the user of some condition in a program, where that condition (normally) doesn’t warrant raising an exception and terminating the program.

如何处理，要看这个warning和它是否影响你的程序执行，你可以:

• 抑制警告
  • 参见 Temporarily Suppressing Warnings
• 将警告视为错误
  • 参见 How do I catch a numpy warning like it's an exception (not just for testing)?