CODEWITHSUNDEEP
pythonarraysnumpynumberscompare
B612
B612

Reputation: 53

How compare inside an array?

I have this array: 

array([[4798, 1369, 4139, ..., 2910, 2901,    0],
       [   2, 3161, 1343, ..., 2085, 2065, 2399],
       [ 546,  506,  614, ..., 1874, 1859, 4799],
       [   0,  634, 1341, ..., 1886, 1871, 4799]], dtype=int64)

And I can see that there are two equal number (4799) in the last rows. That I want is make a function that compare the adjacent rows and return an array or list with the equal numbers. I don't know how to do it. I try do it with a for but when I have an largest array this method is very slow.

Upvotes: 0

Views: 100

Answers (2)

idonthaveanyplan
idonthaveanyplan

Reputation: 56

I make the following assumptions to answer your question:

  • you don't need to detect how often a value is in row A and row B
  • you want to compare one row with ONE other row, row for row
  • the order of the numbers inside of the both rows you compare doesnt matter (if the order does matter, you should use numpy.equal())

First, use the set() function for the both rows you want to compare. Then use the intersection operator & on both sets. It returns a new set with elements that are common to both sets Python Docs - 8.7. sets.

for i in range(len(array)):
    # Start at i+1 to prevent double comparison of row pairs
    for j in range(i+1, len(array)):  
        set_a = set(array[i])
        set_b = set(array[j])
        intersection = set_a & set_b

Upvotes: 1

Nik P
Nik P

Reputation: 224

Let's say a is your array. Then try the following which may help,

def equalnumbers(a):
    row,col = np.where((a[1:,:]-a[:-1,:]) ==0)
    return a[row,col]

Trial example:

Note the array I use has two numbers that have an equal in their adjacent row. There's also the number 0 twice but in non adjacent rows!

import numpy as np
a = np.array([[4798, 1369, 4139, 2910, 2901, 0],
             [4798, 3161, 1343, 2085, 2065, 2399],
             [546,  506,  614, 1874, 1859, 4799],
             [0,  634, 1341, 1886, 1871, 4799]]) 

equalnumbers(a)

Output:

array([4798, 4799])

The function has only returned the equal numbers from adjacent rows and not the 0 which is present in the non adjacent rows. Hope that helps.

Upvotes: 2

Related Questions