NumPy, сортировка массива

    unique,pos = np.unique(A,return_inverse=True)
    counts = np.bincount(pos)                     
    maxsort = counts.argsort()[::-1]
    sorted = ((unique[maxsort],counts[maxsort]))

import numpy as np
arr = [604, 546, 546, 75, 14, 8, 6, 2, 1, 0, 0]
print("unique", np.unique(arr))
counts = np.bincount(arr)
print("bincount", counts)
n = len(counts)
print("n", n)
r = np.arange(n)
print(r[counts>1])
print(r[counts==1])

import numpy as np
arr = [23, 10, 2, 10, 2, 2, 14]
counts = np.bincount(arr)
maxsort = counts.argsort()[::-1]
sorted_arr = ((maxsort,counts[maxsort]))

np.sort(a.view('i4,i4'), order=['f1'], axis=0).view(np.int)

print(r[counts>1], counts[r[counts>1]])
print(r[counts==1], counts[r[counts==1]])

print(r[counts>1], counts[r[counts>1]])
print(r[counts==1], counts[r[counts==1]])

np.sort(a.view('i4,i4'), order=['f1'], axis=0).view(np.int)

np.sort(a.view('i4,i4'), order=['f1'], axis=0).view(np.int)

import numpy as np
arr = [23, 10, 2, 10, 2, 2, 14]
counts = np.bincount(arr)
maxsort = counts.argsort()[::-1]
sorted_arr = ((maxsort,counts[maxsort]))

from collections import Counter
arr = [23, 10, 2, 10, 2, 2, 14]
d = dict(sorted(Counter(arr).items(), key=lambda x: x[1], reverse=True))
print(*d.keys())
print(*d.values())

from collections import Counter
from time import perf_counter
arr = [23, 10, 2, 10, 2, 2, 14]
x = 0
t1_start = perf_counter() 
while x < 100000:
    x+=1
    d = dict(sorted(Counter(arr).items(), key=lambda x: x[1], reverse=True))
t1_stop = perf_counter()
print(round(t1_stop-t1_start, 2), 's')

from time import perf_counter
import numpy as np
arr = [23, 10, 2, 10, 2, 2, 14]
 
 
x = 0
t1_start = perf_counter() 
while x < 100000:
    x+=1
    counts = np.bincount(arr)
    maxsort = counts.argsort()[::-1]
    sorted_arr = ((maxsort,counts[maxsort]))
 
t1_stop = perf_counter()
print(round(t1_stop-t1_start, 2), 's')