Introduction to NumPy

# CHAPTER 10

Introduction to NumPy #

1. Chapter Introduction #

2. What is NumPy? #

NumPy is an open-source Python library used for working with arrays. It also contains functions for linear algebra, fourier transform, and matrices.

The core feature of NumPy is the ndarray (N-dimensional array). It looks like a list, but under the hood, it is implemented in optimized C code.

3. Installing and Importing #

If you are using Anaconda, NumPy is already installed. Otherwise, run !pip install numpy.

# 'np' is the universal industry standard alias
import numpy as np

4. Creating NumPy Arrays #

You can create an array by passing a standard Python list into np.array().

import numpy as np

# Create from a list
my_list = [1, 2, 3, 4, 5]
my_array = np.array(my_list)

print(my_array)         # [1 2 3 4 5]
print(type(my_array))   # <class 'numpy.ndarray'>

The Data Type Restriction: Unlike Python lists, NumPy arrays are *homogeneous*. This means every item in the array MUST be the exact same data type.

# If you mix types, NumPy converts them all to strings!
mixed_list = [1, 2, "Apple"]
mixed_array = np.array(mixed_list)

print(mixed_array) # ['1' '2' 'Apple']

5. Array Creation Functions #

NumPy has built-in functions to quickly generate large arrays without typing them out.

# 1. Array of zeros (useful for initializing empty matrices)
zeros = np.zeros(5)
print(zeros) # [0. 0. 0. 0. 0.]

# 2. Array of ones
ones = np.ones(3)
print(ones) # [1. 1. 1.]

# 3. Arange (like Python's range(), but returns an array)
# np.arange(start, stop, step)
seq = np.arange(0, 10, 2)
print(seq) # [0 2 4 6 8]

# 4. Linspace (Evenly spaced numbers over a specified interval)
lin = np.linspace(0, 1, 5)
print(lin) # [0.   0.25 0.5  0.75 1.  ]

6. Array Properties #

You must understand the shape of your data. NumPy provides attributes to inspect arrays.

# Let's create a 2D Array (A Matrix)
matrix = np.array([
    [1, 2, 3],
    [4, 5, 6]
])

print("Dimensions:", matrix.ndim)  # Output: 2
print("Shape:", matrix.shape)      # Output: (2, 3) -> 2 rows, 3 columns
print("Total Items:", matrix.size) # Output: 6
print("Data Type:", matrix.dtype)  # Output: int64 (or int32)

7. Why NumPy is Faster (A Preview of Vectorization) #

To multiply every number in a Python list by 2, you need a for loop. In NumPy, you apply the math directly to the array.

prices = np.array([10, 20, 30])

# Multiplies every element by 2 instantly, in C code.
doubled = prices * 2 

print(doubled) # [20 40 60]

8. Common Mistakes #

• Forgetting the brackets for 2D arrays: When making a matrix, people often write np.array([1, 2], [3, 4]). This is an error. It must be a list of lists: np.array([ [1, 2], [3, 4] ]). Note the extra set of outer brackets.

• Using Python math module on arrays: If you want the square root of an array, math.sqrt(my_array) will crash. You must use NumPy's equivalent: np.sqrt(my_array).

9. MCQs #

10. Interview Questions #

• Q: Explain the difference between a Python List and a NumPy Array. Why is the NumPy array preferred for data science?

• Q: What does it mean that NumPy arrays are "homogeneous"?

11. Summary #

12. Next Chapter Recommendation #