Python Argmin

Introduction

The argmin function is defined as numpy.argmin(). This function returns the index of the minimum value or element from a Numpy array in a specific axis. An array is taken as the input by this function and gives output as the minimum element index.

The input array can be specified as a single-dimensional array or a multi-dimensional array.

By using argmin function we can also do array sorting.

About Numpy

The Numpy is a package in Python which is generally used for processing arrays.

The Numpy can be easily installed via pip with the following command   

pip install numpy

Import Numpy

It can easily import with the Import keyword

Syntax:

import numpy as np

Argmin installation

Argmin cab be installed via PIP with below command.

pip install argmin

Syntax of Numpy argmin:

numpy.argmin(a, axis=None, out=None)

Parameters :

array: It is the input array for which the minimum element index has to be determined.

axis: This is an optional parameter that takes an integer value. By default, It is none if the value is not defined.

out:  If this parameter is defined, then the result provided by the argmin function is stored in this array. The size of the array is specified in a manner by which it is appropriate with the returned shape of the array.

Return:  index of the minimum value or element from a Numpy array in a specific axis.

Example:

import numpy as np

# Working on 1D array

array = np.arange(8)

print("INPUT ARRAY : \n", array)

# This returns minimum minimum element indices

print("\nIndex of min element : ", np.argmin(array, axis=0))

Output:

[0 1 2 3 4 5 6 7 8 9]
Index of min element : 0

For multidimensional array

The argmin() function can also be used for calculate the minimum value for a multidimensional array. The result will be computed with the axis of the multidimensional array.

Example:

import numpy as np

 # Working on 2D array

array =  np.random.randint(16, size=(4, 4))
print("INPUT ARRAY : \n", array)
print("\nIndices of min element : ", np.argmin(array, axis = 0))

Output:

INPUT ARRAY :
[[ 9 11 10  0]
[10  6  0 11]
[ 6  4 15  2]
[12 12 10  1]]
Indices of min element :  [2 2 1 0]

With the different axis values

Now, we are trying to change the axis values of the argmin function to check how the output will be changed.

Here, we will take axis =1.

Example:

import numpy as np

# Working on 2D array

array =  np.random.randint(16, size=(4, 4))

print("INPUT ARRAY : \n", array)

print("\nIndices of min element : ", np.argmin(array, axis = 1))

Output:

Indices of min element :  [3 2 3 3]

Using argmin with Condition

An argmin() function can also be applied for a presented array where the items satisfy a particular condition and, the Masked array should be used in Numpy for this purpose. Where the array items have invalid entries those arrays are known as Masked arrays.

“When” the items don't satisfy the presented condition they will not be considered, and we have to mask them. For the rest of the items, the minimum value indices will be returned.

Example:

import numpy as np

array = np.array([10,7,3,11,2,8,12,9])

array = np.ma.MaskedArray(array, array<9)

print(array)

Output:

[10 -- -- 11 -- -- 12 9]

Using argmin with Matrix

The argmin function can also be used with matrix to find the index of minimum elements.

Syntax:

numpy.matrix.argmin 

Example:

import numpy as np

matrix = np.matrix(np.arange(25).reshape((5,5)))

print(matrix)

print("\nIndices of min element : ", matrix.argmin())

Output:

[[ 0  1  2  3  4]

[ 5  6  7  8  9]

[10 11 12 13 14]

[15 16 17 18 19]

[20 21 22 23 24]]

Indices of min element :  0

Conclusion

In the above article, we have learned about the Python argmin() function and understood how to use the argmin() function with different methods.