沿动态指定的轴切片一个 numpy 数组

问题描述

我想沿特定轴动态切片一个 numpy 数组.鉴于此:

轴 = 2开始 = 5结束 = 10

我想达到与此相同的结果:

# m 是一些矩阵m[:,:,5:10]

使用这样的东西:

slc = tuple(:,) * len(m.shape)slc[轴] = 切片(开始，结束)米[slc]

但是 : 值不能放在元组中，所以我不知道如何构建切片.

解决方案

我认为一种方法是使用 slice(None):

>>>m = np.arange(2*3*5).reshape((2,3,5))>>>轴，开始，结束 = 2, 1, 3>>>目标 = m[:, :, 1:3]>>>目标数组([[[ 1, 2],[6, 7],[11, 12]],[[16, 17],[21, 22],[26, 27]]])>>>slc = [切片(无)] * len(m.shape)>>>slc[轴] = 切片(开始，结束)>>>np.allclose(m[slc], 目标)真的

我有一种模糊的感觉，我以前为此使用过一个函数，但现在似乎找不到了..

I would like to dynamically slice a numpy array along a specific axis. Given this:

axis = 2
start = 5
end = 10

I want to achieve the same result as this:

# m is some matrix
m[:,:,5:10]

Using something like this:

slc = tuple(:,) * len(m.shape)
slc[axis] = slice(start,end)
m[slc]

But the : values can't be put in a tuple, so I can't figure out how to build the slice.

解决方案

I think one way would be to use slice(None):

>>> m = np.arange(2*3*5).reshape((2,3,5))
>>> axis, start, end = 2, 1, 3
>>> target = m[:, :, 1:3]
>>> target
array([[[ 1,  2],
        [ 6,  7],
        [11, 12]],

       [[16, 17],
        [21, 22],
        [26, 27]]])
>>> slc = [slice(None)] * len(m.shape)
>>> slc[axis] = slice(start, end)
>>> np.allclose(m[slc], target)
True

I have a vague feeling I've used a function for this before, but I can't seem to find it now..

这篇关于沿动态指定的轴切片一个 numpy 数组的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！