Finding the N highest values of a vector

Question

How do I find the N highest values of an vector? Let´s say I´ve got a vector:

arr = np.array([12.5, 13.6, −9.1, 17.5, 15.3, 10.5])

And N = 3

the output should then be:

np.array([13.6,17.5,15.3])

So the 3 highest values of arr are kept in the same order.

if N=4 the output should be:

np.array([12.5,13.6,17.5,15.3])

Clone the array. Order the cloned array. Remove all elements from nth position. Remove in order the elements from original array that arent on new array. — Andre Motta, Aug 10 '18 at 13:42
Is it important that the elements of the resultant array are in the order they were in in the previous array? — Patrick Haugh, Aug 10 '18 at 13:43
I suppose so @PatrickHaugh if not he just need to order it and remove the elements from the nth position on fowards — Andre Motta, Aug 10 '18 at 13:43

score 1 · Answer 1 · answered Aug 10 '18 at 13:45

You might like heapq.nlargest() method. It allows you to take n largest elements of any collections using linear time. I.e. you don't need to sort your original collection spending O(n*log n) time.

import heapq

n = 4
print(np.array(heapq.nlargest(n, arr)))

Finding the N highest values of a vector

1 Answers1