按多个轴对2D numpy数组排序 [英] Sorting a 2D numpy array by multiple axes

问题描述

我有一个形状为(N，2)的2D numpy数组，其中包含N个点(x和y坐标).例如:

I have a 2D numpy array of shape (N,2) which is holding N points (x and y coordinates). For example:

array([[3, 2],
       [6, 2],
       [3, 6],
       [3, 4],
       [5, 3]])

我想对它进行排序，以使我的点按x坐标排序，然后在x坐标相同的情况下按y排序.因此，上面的数组应如下所示:

I'd like to sort it such that my points are ordered by x-coordinate, and then by y in cases where the x coordinate is the same. So the array above should look like this:

array([[3, 2],
       [3, 4],
       [3, 6],
       [5, 3],
       [6, 2]])

如果这是一个普通的Python列表，我将简单地定义一个比较器来执行我想要的操作，但是据我所知，numpy的sort函数不接受用户定义的比较器.有什么想法吗?

If this was a normal Python list, I would simply define a comparator to do what I want, but as far as I can tell, numpy's sort function doesn't accept user-defined comparators. Any ideas?

感谢您的想法！我建立了一个具有1000000个随机整数点的快速测试用例，并对可以运行的基准点进行了基准测试(对不起，目前无法升级numpy).

Thanks for the ideas! I set up a quick test case with 1000000 random integer points, and benchmarked the ones that I could run (sorry, can't upgrade numpy at the moment).

Mine:   4.078 secs 
mtrw:   7.046 secs
unutbu: 0.453 secs

推荐答案

使用

---

如果a是C_CONTIGUOUS，则

a.ravel()返回一个视图.如果是这样的话 @ars的方法(通过使用ravel而不是flatten进行了稍微修改)产生了对a进行排序的好方法就地:

---

a.ravel() returns a view if a is C_CONTIGUOUS. If that is true, @ars's method, slightly modifed by using ravel instead of flatten, yields a nice way to sort a in-place:

a = np.array([(3, 2), (6, 2), (3, 6), (3, 4), (5, 3)])
dt = [('col1', a.dtype),('col2', a.dtype)]
assert a.flags['C_CONTIGUOUS']
b = a.ravel().view(dt)
b.sort(order=['col1','col2'])

由于b是a的视图，因此对b进行排序也对a进行排序:

Since b is a view of a, sorting b sorts a as well:

print(a)
# [[3 2]
#  [3 4]
#  [3 6]
#  [5 3]
#  [6 2]]

这篇关于按多个轴对2D numpy数组排序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IT屋！