二维Python数组排序:对第二个值相同的第一个数组进行排序

Question

我目前正在处理坐标，不知何故，我需要对具有相同y坐标的x坐标按升序排序，而不触及y坐标。

下面是示例输入:

x = [213,212,213,214,215,216,215,216,217,216,218,217,219,220,219]
y = [352,333,332,330,328,327,327,326,325,325,324,324,323,322,322]
xy = np.array([x,y])

下面是期望的输出:

x = [213,212,213,214,215,215,216,216,216,217,217,218,219,219,220]
y = [352,333,332,330,328,327,327,326,325,325,324,324,323,322,322]

注意:我一直在寻找这个问题，但我似乎找不到直接的答案。谢谢你。

Answer 1

你可以先做一个pair的列表，然后按第一个坐标排序，然后把它做成numpy数组

>>> x = [213,212,213,214,215,216,215,216,217,216,218,217,219,220,219]
>>> y = [352,333,332,330,328,327,327,326,325,325,324,324,323,322,322]
>>> xy=sorted(zip(x,y),key=lambda pair:pair[0])
>>> xy
[(212, 333), (213, 352), (213, 332), (214, 330), (215, 328), (215, 327), (216, 327), (216, 326), (216, 325), (217, 325), (217, 324), (218, 324), (219, 323), (219, 322), (220, 322)]
>>> nxy = np.array(xy).T
>>> nxy
array([[212, 213, 213, 214, 215, 215, 216, 216, 216, 217, 217, 218, 219,
219, 220],
[333, 352, 332, 330, 328, 327, 327, 326, 325, 325, 324, 324, 323,
322, 322]])
>>> nxy
array([[212, 213, 213, 214, 215, 215, 216, 216, 216, 217, 217, 218, 219,
219, 220],
[333, 352, 332, 330, 328, 327, 327, 326, 325, 325, 324, 324, 323,
322, 322]])
>>> 

经过一些好的旧的试验和错误找到一种在numpy中对它们排序的方法

>>> x = [0, 500, 100, 300, 200, 800, 400, 600, 700, 900]
>>> y = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
>>> xy= np.array([x,y])
>>> xy
array([[  0, 500, 100, 300, 200, 800, 400, 600, 700, 900],
[  0,   1,   2,   3,   4,   5,   6,   7,   8,   9]])
>>> np.argsort(xy[0]) #get an array with the index that correspond to the elements in a sorted fashion 
array([0, 2, 4, 3, 6, 1, 7, 8, 5, 9], dtype=int64)
>>> xy[:,np.argsort(xy[0])] #ask for the elements in the aforementioned order, while preserving their pairing
array([[  0, 100, 200, 300, 400, 500, 600, 700, 800, 900],
[  0,   2,   4,   3,   6,   1,   7,   8,   5,   9]])
>>>