numpy.select() function – Python

The numpy.select() function is used to construct an array by selecting elements from a list of choices based on multiple conditions. It is particularly useful when dealing with conditional replacements or transformations in NumPy arrays.

Example:

import numpy as np
arr = np.array([10, 20, 30, 40])
conditions = [arr < 20, arr > 30]
choices = [100, 200]
result = np.select(conditions, choices, default=0)
print(result)

Syntax:

numpy.select(condlist, choicelist, default=0)

Parameters:

• condlist : list of bool ndarrays
  A list of boolean NumPy arrays that determine from which array in choicelist the output elements are selected. If multiple conditions are True, the first one encountered is used.
• choicelist : list of ndarrays
  A list of arrays from which the output elements are chosen. It must have the same length as condlist.
• default : scalar, optional (default=0)
  The value inserted in the output array when none of the conditions are met.

Return Value:

• ndarray : An array with elements chosen from choicelist based on the conditions in condlist.

Code Implementation

1. Basic Usage of numpy.select()

Here:

• If arr < 3, the corresponding element is taken from arr.
• If arr > 4, the corresponding element is taken from arr**3.
• Otherwise, the default value 0 is used.

import numpy as np
arr = np.arange(8)
condlist = [arr < 3, arr > 4]
choicelist = [arr, arr**3]
result = np.select(condlist, choicelist)
print(result)

Output :

[ 0 1 2 0 0 125 216 343]

2. Using a Different Default Value

Here:

• Values where arr < 4 are taken from arr.
• Values where arr > 6 are taken from arr**2.
• All other values are replaced with -1 (instead of 0).

arr = np.arange(8)
condlist = [arr < 4, arr > 6]
choicelist = [arr, arr**2]

# Custom default value (e.g., -1)
result = np.select(condlist, choicelist, default=-1)
print(result)

Output:

0 1 2 3 -1 -1 -1 49]

3. Handling Multiple Conditions

Here :

• If arr is even (arr % 2 == 0), it is multiplied by 10.
• If arr is divisible by 3 (arr % 3 == 0), it is negated.
• If neither condition is met, the default value 100 is used.

arr = np.arange(10)
condlist = [arr % 2 == 0, arr % 3 == 0]
choicelist = [arr * 10, arr * -1]

result = np.select(condlist, choicelist, default=100)
print(result)

Output:

[ 0 100 20 -3 40 100 60 100 80 -9]

Why Use numpy.select()?

• More flexible than numpy.where() when dealing with multiple conditions.
• Helps avoid complex nested if-else conditions in array transformations.
• Efficient and concise for applying different transformations to an array based on conditions.

Comparison with numpy.where()

The numpy.select() function is a powerful tool for conditional selection and transformation of array elements. It is especially useful when handling multiple conditions efficiently in a structured way. Mastering its usage will help simplify complex array operations in Python.

GeeksforGeeks

Improve

Article Tags :

• Machine Learning
• Numpy
• python
• Python numpy-arrayManipulation
• Python-numpy

+1 More

Practice Tags :

• Machine Learning
• python