Basics of NumPy Arrays

NumPy stands for Numerical Python and is used for handling large, multi-dimensional arrays and matrices. Unlike Python's built-in lists NumPy arrays provide efficient storage and faster processing for numerical and scientific computations. It offers functions for linear algebra and random number generation making it important for data science and machine learning.

Types of Array

1. One Dimensional Array: stores elements in a single row. It is the simplest type of NumPy array and is commonly used to represent a sequence of values.

Example: The following example creates a one-dimensional NumPy array from a Python list.

import numpy as np

a = [1, 2, 3, 4]
arr = np.array(a)

print("List: ", a)
print("Numpy Array:", arr)
print(type(a))
print(type(arr))

List:  [1, 2, 3, 4]
Numpy Array: [1 2 3 4]
<class 'list'>
<class 'numpy.ndarray'>

2. Multi-Dimensional Array: stores data in two or more dimensions. The most common type is a two-dimensional array, which organizes data into rows and columns.

Example: The following example creates a two-dimensional NumPy array from multiple Python lists.

import numpy as np

l1 = [1, 2, 3, 4]
l2 = [5, 6, 7, 8]
l3 = [9, 10, 11, 12]
arr = np.array([l1, l2, l3])
print(arr)

[[ 1  2  3  4]
 [ 5  6  7  8]
 [ 9 10 11 12]]

Explanation:

• Each inner list represents a row in the array.
• np.array() combines the lists into a two-dimensional NumPy array.

Parameters of a Numpy Array

1. Axis: represents a dimension of an array.

Axis 0 -> Rows (first dimension)
Axis 1 -> Columns (second dimension)
Axis 2 -> Third dimension in a 3D array

For example, a two-dimensional array has Axis 0 and Axis 1.

2. Shape: indicates the number of elements along each dimension. It is returned as a tuple.

import numpy as np
arr = np.array([ [0, 4, 2],
                 [3, 4, 5],
                 [23, 4, 5],
                 [2, 34, 5],
                 [5, 6, 7] ])
print(arr.shape)

(5, 3)

Explanation: array contains 5 rows and 3 columns. Therefore, arr.shape returns (5, 3)

3. Rank: number of dimensions (axes) it has.

• A one-dimensional array has rank 1.
• A two-dimensional array has rank 2.
• A three-dimensional array has rank 3.

np.array([1, 2, 3])      # Rank 1
np.array([[1, 2], [3, 4]])  # Rank 2

4. Data Type (dtype): property specifies the type of data stored in an array, such as integers, floating-point numbers, or strings.

import numpy as np

arr1 = np.array([[0, 4, 2]])
arr2 = np.array([0.2, 0.4, 2.4])

print("Data type of array 1:", arr1.dtype)
print("Data type of array 2:", arr2.dtype)

Data type of array 1: int64
Data type of array 2: float64

Different Ways of Creating Numpy Array

1. numpy.array(): creates a NumPy array from Python sequences such as lists or tuples.

import numpy as np
arr = np.array([3, 4, 5, 5])
print(arr)

[3 4 5 5]

2. numpy.fromiter(): creates a one-dimensional array from an iterable object.

import numpy as np
text = "Geeksforgeeks"
arr = np.fromiter(text, dtype="U2")
print(arr)

['G' 'e' 'e' 'k' 's' 'f' 'o' 'r' 'g' 'e' 'e' 'k' 's']

3. numpy.arange(): returns evenly spaced values within a specified range.

import numpy as np
arr = np.arange(1, 20, 2, dtype=np.float32)
print(arr)

[ 1.  3.  5.  7.  9. 11. 13. 15. 17. 19.]

4. numpy.linspace(): returns a specified number of evenly spaced values between two limits.

import numpy as np
arr = np.linspace(3.5, 10, 3, dtype=np.int32)
print(arr)

[ 3  6 10]

5. numpy.empty(): creates an array of a given shape without initializing its values.

import numpy as np
arr = np.empty((4, 3), dtype=np.int32)
print(arr)

[[   3130309      21840          0]
 [         0  540950078 1920216673]
 [1886613089  677737327 1918962217]
 [ 679043442  539767131  857746482]]

6. numpy.ones(): creates an array filled with ones.

import numpy as np
arr = np.ones((4, 3), dtype=np.int32)
print(arr)

[[1 1 1]
 [1 1 1]
 [1 1 1]
 [1 1 1]]

7. numpy.zeros(): creates an array filled with zeros.

import numpy as np
arr = np.zeros((4, 3), dtype=np.int32)
print(arr)