Sort array based on value and create new array

Question

Imagine a two dimensional array:

a = np.array([[1,1],[1, 0],[0, 0],[0, 0],[0, 0],[1, 1],[1, 1],[0, 1]])

I want to sort the array based on its first value like:

[[1,1],[1, 1],[1, 1],[1, 0],[0, 1],[0, 0],[0, 0],[0, 0]]

If I am simply going with a .sort() like:

a[::-1].sort(axis=0)

and the returned array looks like:

array([[1, 1],
       [1, 1],
       [1, **1**],
       [**1**, 1],
       [0, 0],
       [0, 0],
       [0, 0],
       [0, 0]])

As you can see the bold 1 used to be a zero. Why is the function flipping around my numbers? I searched the internet and haven't found any answers.

score 1 · Answer 1 · answered Jan 29 '21 at 22:35

the sort you are doing is a sort of all columns in the array independently of each other, from the first example on this page https://numpy.org/doc/stable/reference/generated/numpy.sort.html

>>> a = np.array([[1,4],[3,1]])
>>> np.sort(a)                # sort along the last axis
array([[1, 4],
       [1, 3]])
>>> np.sort(a, axis=None)     # sort the flattened array
array([1, 1, 3, 4])
>>> np.sort(a, axis=0)        # sort along the first axis
array([[1, 1],
       [3, 4]])

also see this answer to sort the rows based on a single column: Sorting arrays in NumPy by column

score 1 · Accepted Answer · answered Jan 29 '21 at 22:36

The problem is that numpy sort when you pass axis=0 is sorting each column independently (see examples on doc page). If you want to sort rows, then you can use sorted instead:

np.array(sorted(a, key=lambda x: x.tolist(), reverse=True))

In your case the result is

[[1 1]
 [1 1]
 [1 1]
 [1 0]
 [0 1]
 [0 0]
 [0 0]
 [0 0]]

score 1 · Answer 3 · answered Jan 29 '21 at 22:37

You can use np.lexsort, and pass the two columns independently, then reverse the order. lexsort returns the sorted indices, given the key. You need to put first column second, because the primary key in lexsort is the last column:

>>> a[np.lexsort((a[:,1], a[:,0]))][::-1]
 
array([[1, 1],
       [1, 1],
       [1, 1],
       [1, 0],
       [0, 1],
       [0, 0],
       [0, 0],
       [0, 0]])

Here is the output of np.lexsort on your data:

                 key1     key2 (primary)
>>> np.lexsort((a[:,1], a[:,0]))
 array([2, 3, 4, 7, 1, 0, 5, 6], dtype=int64)

Sort array based on value and create new array

3 Answers3