python中有没有直接对多维数组排序的方法？

Question

如何按第一列的降序排列如下数组：

dl1 = numpy.array([[ 0.02598003,1.],
                   [ 0.00730082,2.],
                   [ 0.05471569,3.],
                   [ 0.02599167,4.],
                   [ 0.0544947 ,5.],
                   [ 0.00753346,6.]])

网上其他地方说直接dl1.sort()会默认按第一列排序，但好像不行

Answer 1

sorted(dl1, key=lambda x: x[0])

Answer 2

>>> a=np.array([[ 0.02598003,1.],
               [ 0.00730082,2.],
               [ 0.05471569,3.],
               [ 0.02599167,4.],
               [ 0.0544947 ,5.],
               [ 0.00753346,6.]])
>>> a.sort(0)
>>> a
array([[ 0.00730082,  1.        ],
       [ 0.00753346,  2.        ],
       [ 0.02598003,  3.        ],
       [ 0.02599167,  4.        ],
       [ 0.0544947 ,  5.        ],
       [ 0.05471569,  6.        ]])
>>> 

---

np.sort 是把各维分别排序的

如果你是要二维组的联合排序，要用np.argsort方法

>>> a=np.array([[ 0.02598003,1.],
               [ 0.00730082,2.],
               [ 0.05471569,3.],
               [ 0.02599167,4.],
               [ 0.0544947 ,5.],
               [ 0.00753346,6.]])

>>> a[a.argsort(0)[:,0]]
array([[ 0.00730082,  2.        ],
       [ 0.00753346,  6.        ],
       [ 0.02598003,  1.        ],
       [ 0.02599167,  4.        ],
       [ 0.0544947 ,  5.        ],
       [ 0.05471569,  3.        ]])
>>> 

如果数据很多的话，用python内部的 sorted会降低效率

Answer 3

In [1]: lst= [[0.00730082, 2.0],
   ...:  [0.05471569, 3.0],
   ...:  [0.02599167, 4.0],
   ...:  [0.0544947, 5.0],
   ...:  [0.00753346, 6.0]]
   ...:

In [2]: sorted(lst, key=lambda x: x[0])
Out[2]:
[[0.00730082, 2.0],
 [0.00753346, 6.0],
 [0.02599167, 4.0],
 [0.0544947, 5.0],
 [0.05471569, 3.0]]

Answer 4

dl1.sort(axis=0)

ndarray.sort的关键字参数axis就是用来按照某列排序

axis : int, optional

Axis along which to sort. Default is -1, which means sort along the last axis.