将2D numpy矩阵转换为2D numpy数组

Question

Certain functions in Numpy return a 2d matrix as output. but I want them to be in 2d array form. What is the most efficient (memory and cpu) way to convert a 2d matrix to a 2d array?

Answer 1

Note that a numpy.matrix is already an ndarray subclass, and nothing more than a specialized 2D array. Hence you're most likely quite alright without converting your matrix to an explicit numpy.array unless you have a particular reason to do so, perhaps the additional generality of a Numpy array.

Should this be the case, you can convert your matrix to an array with numpy.asarray() . It's important you use this method and not numpy.asanyarray() in your case as with numpy.asanyarray() allows subclasses of ndarray to pass through, as your matrix would.