Python：NumPy排序数组

Question

I am currently trying to sort a numpy array, but i am running into a difficulty

The array that i want to sort is the following:

mat = np.array([0.05170475 0.07367926 0.05741241 0.34870369 0.19990381 0.26859608])

Now the difficult part here, is that i want to sort the array, but also keep the indexes at the same time.

For example without using numpy , i would have used

mat = list(enumerate(mat))    # gives [(0, 0.05170474575702143), (1, 0.07367926270375554), (2, 0.05741241249643288), (3, 0.3487036852148175), (4, 0.19990381197331886), (5, 0.2685960818546567)]
mat.sort(reverse = True, key = lambda ×: ×[1])    # gives [(3, 0.3487036852148175), (5, 0.2685960818546567), (4, 0.19990381197331886), (1, 0.07367926270375554), (2, 0.05741241249643288), (0, 0.05170474575702143)]

However, since i am using numpy , i was wondering if there was maybe a numpy function that can do all of that. I was able to use np.sort and np.argsort to sort the indexes and the values individually, but i wasn't able to do both at the same time...

Answer 1

indices = np.argsort(mat)[::-1]
np.hstack((indices[:, np.newaxis], mat[indices][:, np.newaxis]))

array([[3.        , 0.34870369],
       [5.        , 0.26859608],
       [4.        , 0.19990381],
       [1.        , 0.07367926],
       [2.        , 0.05741241],
       [0.        , 0.05170475]])