How to create an empty and a full NumPy array

Creating arrays is a basic operation in NumPy. Two commonly used types are:

• Empty array: This array isn’t initialized with any specific values. It’s like a blank page, ready to be filled with data later. However, it will contain random leftover values in memory until you update it.
• Full array: This is an array where all the elements are set to the same specific value right from the start. It’s like a sheet filled with one number everywhere.

Now let’s understand how to create an empty array and a full array in detail.

Creating Empty Array

Creating an empty array is useful when you need a placeholder for future data that will be populated later. It allocates space without initializing it, which can be efficient in terms of performance.

• Use np.empty() function.
• Specify the shape of the array as a tuple.
• Optionally, define the data type using the dtype parameter.

Syntax:

numpy.empty(shape, dtype=float, order='C')

Parameters:

• shape (tuple of int): Shape of the new array
• dtype (data-type, optional): data type of the array (default is float).
• order ({'C', 'F'}, optional): Memory layout. 'C' means row-major (C-style), 'F' means column-major (Fortran-style).

Example: This code creates a 3x4 empty array using np.empty().

import numpy as np
empArr = np.empty((3, 4))
print(empArr)

[[4.68337834e-310 0.00000000e+000 0.00000000e+000 0.00000000e+000]
 [0.00000000e+000 0.00000000e+000 0.00000000e+000 0.00000000e+000]
 [0.00000000e+000 0.00000000e+000 0.00000000e+000 0.00000000e+000]...

Explanation:

• np.empty((3, 4)): Creates an array with shape (3,4).
• The values are uninitialized garbage values (whatever is present in memory).
• This is faster than creating an array filled with zeros or ones, but you should assign values before using it.

Creating Full Array

A full array is ideal when you need an array initialized with a specific value, such as zeros or ones, which is common in many mathematical computations.

• Use np.full() function.
• Pass the desired shape and fill value.
• Optionally, specify the data type.

Syntax:

numpy.full(shape, fill_value, dtype=None, order='C')

Parameters:

• shape (tuple of int): Shape of the new array.
• fill_value (scalar): The constant value to fill the array with.
• dtype (data-type, optional): data type of the array (inferred if not given).
• order ({'C', 'F'}, optional): Memory layout.

Example: This code creates a 3x4 array filled with the value 5 using np.full().

import numpy as np
fullArr = np.full((3, 4), 5)
print(fullArr)

[[5 5 5 5]
 [5 5 5 5]
 [5 5 5 5]]

Explanation:

• np.full((3, 4), 5): Creates an array of shape (3,4) where every element is 5.
• Unlike np.empty(), this array is fully initialized with the given value.
• Useful when you need a matrix with constant values for calculations.