Sort Array in NumPy by the n-1th Column

In this example, we will see how to sort an array in Numpy by the (N-1)th column.

Sort an array by (n-1) th column using argsort()

Example

Let us see the first example to sort an array by (n-1)th column ?

import numpy as np

# Creat a Numpy Array
a = np.array([[9, 2, 3],
   [4, 5, 6],
   [7, 0, 5]])

print("Array = \n",a)

# The value of n
n = 3

# Sort by n-1 column
print("\nSort by n-1 th column = \n",a[a[:, (n-1)].argsort()])

Output

Array = 
 [[9 2 3]
 [4 5 6]
 [7 0 5]]

Sort by n-1 th column = 
 [[9 2 3]
 [7 0 5]
 [4 5 6]]

Sort an array by (n-1) th column using sorted() and lambda

Example

In this example, we will sort using lambda ?

import numpy as np

# Create a Numpy Array
arr = np.array([[5, 7, 7],
   [2, 3, 9],
   [7, 6, 6]])

print("Array = \n",arr)

# The value of n
n = 3

# Sort by n-1 column using lambda
arr = sorted(arr, key=lambda arr_entry: arr_entry[n-1])

print("\nSort by n-1 th column = \n",arr)

Output

Array = 
 [[5 7 7]
 [2 3 9]
 [7 6 6]]

Sort by n-1 th column = 
 [array([7, 6, 6]), array([5, 7, 7]), array([2, 3, 9])]

Sort an array by (n-1) th column using mergesort

Example

In this example, we will sort using the mergesort() method in Python ?

import numpy as np

# Create a Numpy Array
arr = np.array([[5, 7, 7],
   [2, 3, 9],
   [7, 6, 6]])

print("Array = \n",arr)

# The value of n
n = 3

# Sort by n-1 column using mergesort
arr = arr[arr[:,(n-1)].argsort(kind='mergesort')]
print("\nSort by n-1 th column = \n",arr)

Output

Array = 
 [[5 7 7]
 [2 3 9]
 [7 6 6]]

Sort by n-1 th column = 
 [[7 6 6]
 [5 7 7]
 [2 3 9]]