numpy.nanargmax() in Python (original) (raw)

Last Updated : 19 Sep, 2023

The numpy.nanargmax() function returns indices of the max element of the array in a particular axis ignoring NaNs.
The results cannot be trusted if a slice contains only NaNs and Infs.

Syntax:

numpy.nanargmax(array, axis = None)

Parameters :

array : Input array to work on axis : [int, optional]Along a specified axis like 0 or 1

Return :

Array of indices into the array with same shape as array.shape with the dimension along axis removed.

Code 1 :

Python `

Python Program illustrating

working of nanargmax()

import numpy as geek

Working on 1D array

array = [geek.nan, 4, 2, 3, 1] print("INPUT ARRAY 1 : \n", array)

array2 = geek.array([[geek.nan, 4], [1, 3]])

returning Indices of the max element

as per the indices ingnoring NaN

print("\nIndices of max in array1 : ", geek.nanargmax(array))

Working on 2D array

print("\nINPUT ARRAY 2 : \n", array2) print("\nIndices of max in array2 : ", geek.nanargmax(array2))

print("\nIndices at axis 1 of array2 : ", geek.nanargmax(array2, axis = 1))

`

Output :

INPUT ARRAY 1 : [nan, 4, 2, 3, 1]

Indices of max in array1 : 1

INPUT ARRAY 2 : [[ nan 4.] [ 1. 3.]]

Indices of max in array2 : 1

Indices at axis 1 of array2 : [1 1]

Code 2: Comparing working of argmax and nanargmax

Python `

Python Program illustrating

working of nanargmax()

import numpy as geek

Working on 2D array

array = ( [[ 8, 13, 5, 0], [ 16, geek.nan, 5, 3], [geek.nan, 7, 15, 15], [3, 11, 4, 12]]) print("INPUT ARRAY : \n", array)

returning Indices of the max element

as per the indices

'''
[[ 8 13 5 0] [ 16 2 5 3] [10 7 15 15] [ 3 11 4 12]] ^ ^ ^ ^

'''

print("\nIndices of max using argmax : ", geek.argmax(array, axis = 0)) print("\nIndices of max using nanargmax : : ", geek.nanargmax(array, axis = 0))

`

Output :

INPUT ARRAY : [[8, 13, 5, 0], [16, nan, 5, 3], [nan, 7, 15, 15], [3, 11, 4, 12]]

Indices of max using argmax : [2 1 2 2]

Indices of max using nanargmax : : [1 0 2 2]

Note :
These codes won't run on online IDE's. So please, run them on your systems to explore the working.