2/24/23, 12:00 AM Copy of colab-tutorial.

ipynb - Colaboratory

CS231n Python Tutorial With Google Colab

This tutorial was originally written by Justin Johnson for cs231n. It was adapted as a Jupyter
notebook for cs228 by Volodymyr Kuleshov and Isaac Caswell.

This version has been adapted for Colab by Kevin Zakka for the Spring 2020 edition of cs231n. It
runs Python3 by default.

Introduction

Python is a great general-purpose programming language on its own, but with the help of a few
popular libraries (numpy, scipy, matplotlib) it becomes a powerful environment for scientific
computing.

We expect that many of you will have some experience with Python and numpy; for the rest of you,
this section will serve as a quick crash course both on the Python programming language and on
the use of Python for scientific computing.

Some of you may have previous knowledge in Matlab, in which case we also recommend the
numpy for Matlab users page (https://docs.scipy.org/doc/numpy-dev/user/numpy-for-matlab-
users.html).

In this tutorial, we will cover:

Basic Python: Basic data types (Containers, Lists, Dictionaries, Sets, Tuples), Functions,
Classes
Numpy: Arrays, Array indexing, Datatypes, Array math, Broadcasting
Matplotlib: Plotting, Subplots, Images
IPython: Creating notebooks, Typical workflows

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 1/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

A Brief Note on Python Versions

As of Janurary 1, 2020, Python has officially dropped support for python2 . We'll be using Python 3.7
for this iteration of the course. You can check your Python version at the command line by running
python --version . In Colab, we can enforce the Python version by clicking Runtime -> Change
Runtime Type and selecting python3 . Note that as of April 2020, Colab uses Python 3.6.9 which
should run everything without any errors.

!python --version

Python 3.6.9

Basics of Python

Python is a high-level, dynamically typed multiparadigm programming language. Python code is

often said to be almost like pseudocode, since it allows you to express very powerful ideas in very
few lines of code while being very readable. As an example, here is an implementation of the
classic quicksort algorithm in Python:

def quicksort(arr):
    if len(arr) <= 1:
        return arr
    pivot = arr[len(arr) // 2]
    left = [x for x in arr if x < pivot]
    middle = [x for x in arr if x == pivot]
    right = [x for x in arr if x > pivot]
    return quicksort(left) + middle + quicksort(right)

print(quicksort([3,6,8,10,1,2,1]))

[1, 1, 2, 3, 6, 8, 10]

Basic data types

Numbers

Integers and floats work as you would expect from other languages:

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 2/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

x = 3
print(x, type(x))

3 <class 'int'>
ERROR! Session/line number was not unique in database. History logging moved to new sess

print(x + 1)   # Addition
print(x - 1)   # Subtraction
print(x * 2)   # Multiplication
print(x ** 2)  # Exponentiation

4
2
6
9

x += 1
print(x)
x *= 2
print(x)

9
18

y = 2.5
print(type(y))
print(y, y + 1, y * 2, y ** 2)

<class 'float'>
2.5 3.5 5.0 6.25

Note that unlike many languages, Python does not have unary increment (x++) or decrement (x--)
operators.

Python also has built-in types for long integers and complex numbers; you can find all of the details
in the documentation.

Booleans

Python implements all of the usual operators for Boolean logic, but uses English words rather than
symbols ( && , || , etc.):

t, f = True, False
print(type(t))

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 3/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

<class 'bool'>

Now we let's look at the operations:

print(t and f) # Logical AND;
print(t or f)  # Logical OR;
print(not t)   # Logical NOT;
print(t != f)  # Logical XOR;

False
True
False
True

Strings

hello = 'hello'   # String literals can use single quotes
world = "world"   # or double quotes; it does not matter
print(hello, len(hello))

hello 5

hw = hello + ' ' + world  # String concatenation
print(hw)

hello world

hw12 = '{} {} {}'.format(hello, world, 12)  # string formatting
print(hw12)

hello world 12

String objects have a bunch of useful methods; for example:

s = "hello"
print(s.capitalize())  # Capitalize a string
print(s.upper())       # Convert a string to uppercase; prints "HELLO"
print(s.rjust(7))      # Right-justify a string, padding with spaces
print(s.center(7))     # Center a string, padding with spaces
print(s.replace('l', '(ell)'))  # Replace all instances of one substring with another
print('  world '.strip())  # Strip leading and trailing whitespace

Hello
HELLO
hello

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 4/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

hello
he(ell)(ell)o
world

You can find a list of all string methods in the documentation.

Containers

Python includes several built-in container types: lists, dictionaries, sets, and tuples.

Lists

A list is the Python equivalent of an array, but is resizeable and can contain elements of different
types:

xs = [3, 1, 2]   # Create a list
print(xs, xs[2])
print(xs[-1])     # Negative indices count from the end of the list; prints "2"

[3, 1, 2] 2
2

xs[2] = 'foo'    # Lists can contain elements of different types
print(xs)

[3, 1, 'foo']

xs.append('bar') # Add a new element to the end of the list
print(xs)  

[3, 1, 'foo', 'bar']

x = xs.pop()     # Remove and return the last element of the list
print(x, xs)

bar [3, 1, 'foo']

As usual, you can find all the gory details about lists in the documentation.

Slicing

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 5/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

In addition to accessing list elements one at a time, Python provides concise syntax to access
sublists; this is known as slicing:

nums = list(range(5))    # range is a built-in function that creates a list of integers
print(nums)         # Prints "[0, 1, 2, 3, 4]"
print(nums[2:4])    # Get a slice from index 2 to 4 (exclusive); prints "[2, 3]"
print(nums[2:])     # Get a slice from index 2 to the end; prints "[2, 3, 4]"
print(nums[:2])     # Get a slice from the start to index 2 (exclusive); prints "[0, 1]"
print(nums[:])      # Get a slice of the whole list; prints ["0, 1, 2, 3, 4]"
print(nums[:-1])    # Slice indices can be negative; prints ["0, 1, 2, 3]"
nums[2:4] = [8, 9] # Assign a new sublist to a slice
print(nums)         # Prints "[0, 1, 8, 9, 4]"

[0, 1, 2, 3, 4]
[2, 3]
[2, 3, 4]
[0, 1]
[0, 1, 2, 3, 4]
[0, 1, 2, 3]
[0, 1, 8, 9, 4]

Loops

You can loop over the elements of a list like this:

animals = ['cat', 'dog', 'monkey']
for animal in animals:
    print(animal)

cat
dog
monkey

If you want access to the index of each element within the body of a loop, use the built-in
enumerate function:

animals = ['cat', 'dog', 'monkey']
for idx, animal in enumerate(animals):
    print('#{}: {}'.format(idx + 1, animal))

#1: cat
#2: dog
#3: monkey

List comprehensions:
https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 6/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

When programming, frequently we want to transform one type of data into another. As a simple
example, consider the following code that computes square numbers:

nums = [0, 1, 2, 3, 4]
squares = []
for x in nums:
    squares.append(x ** 2)
print(squares)

[0, 1, 4, 9, 16]

You can make this code simpler using a list comprehension:

nums = [0, 1, 2, 3, 4]
squares = [x ** 2 for x in nums]
print(squares)

[0, 1, 4, 9, 16]

List comprehensions can also contain conditions:

nums = [0, 1, 2, 3, 4]
even_squares = [x ** 2 for x in nums if x % 2 == 0]
print(even_squares)

[0, 4, 16]

Dictionaries

A dictionary stores (key, value) pairs, similar to a Map in Java or an object in Javascript. You can use
it like this:

d = {'cat': 'cute', 'dog': 'furry'}  # Create a new dictionary with some data
print(d['cat'])       # Get an entry from a dictionary; prints "cute"
print('cat' in d)     # Check if a dictionary has a given key; prints "True"

cute
True

d['fish'] = 'wet'    # Set an entry in a dictionary
print(d['fish'])      # Prints "wet"

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 7/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

wet

print(d['monkey'])  # KeyError: 'monkey' not a key of d

---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
<ipython-input-31-78fc9745d9cf> in <module>()
----> 1 print(d['monkey']) # KeyError: 'monkey' not a key of d

KeyError: 'monkey'

SEARCH STACK OVERFLOW

print(d.get('monkey', 'N/A'))  # Get an element with a default; prints "N/A"
print(d.get('fish', 'N/A'))    # Get an element with a default; prints "wet"

N/A
wet

del d['fish']        # Remove an element from a dictionary
print(d.get('fish', 'N/A')) # "fish" is no longer a key; prints "N/A"

N/A

You can find all you need to know about dictionaries in the documentation.

It is easy to iterate over the keys in a dictionary:

d = {'person': 2, 'cat': 4, 'spider': 8}
for animal, legs in d.items():
    print('A {} has {} legs'.format(animal, legs))

A person has 2 legs

A cat has 4 legs
A spider has 8 legs

Dictionary comprehensions: These are similar to list comprehensions, but allow you to easily
construct dictionaries. For example:

nums = [0, 1, 2, 3, 4]
even_num_to_square = {x: x ** 2 for x in nums if x % 2 == 0}
print(even_num_to_square)

{0: 0, 2: 4, 4: 16}

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 8/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

Sets

A set is an unordered collection of distinct elements. As a simple example, consider the following:

animals = {'cat', 'dog'}
print('cat' in animals)   # Check if an element is in a set; prints "True"
print('fish' in animals)  # prints "False"

True
False

animals.add('fish')      # Add an element to a set
print('fish' in animals)
print(len(animals))       # Number of elements in a set;

True
3

animals.add('cat')       # Adding an element that is already in the set does nothing
print(len(animals))       
animals.remove('cat')    # Remove an element from a set
print(len(animals))       

3
2

Loops: Iterating over a set has the same syntax as iterating over a list; however since sets are
unordered, you cannot make assumptions about the order in which you visit the elements of the set:

animals = {'cat', 'dog', 'fish'}
for idx, animal in enumerate(animals):
    print('#{}: {}'.format(idx + 1, animal))

#1: dog
#2: cat
#3: fish

Set comprehensions: Like lists and dictionaries, we can easily construct sets using set
comprehensions:

from math import sqrt
print({int(sqrt(x)) for x in range(30)})

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 9/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

{0, 1, 2, 3, 4, 5}

Tuples

A tuple is an (immutable) ordered list of values. A tuple is in many ways similar to a list; one of the
most important differences is that tuples can be used as keys in dictionaries and as elements of
sets, while lists cannot. Here is a trivial example:

d = {(x, x + 1): x for x in range(10)}  # Create a dictionary with tuple keys
t = (5, 6)       # Create a tuple
print(type(t))
print(d[t])       
print(d[(1, 2)])

<class 'tuple'>
5
1

t[0] = 1

---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-43-c8aeb8cd20ae> in <module>()
----> 1 t[0] = 1

TypeError: 'tuple' object does not support item assignment

SEARCH STACK OVERFLOW

Functions

Python functions are defined using the def keyword. For example:

def sign(x):
    if x > 0:
        return 'positive'
    elif x < 0:
        return 'negative'
    else:
        return 'zero'

for x in [-1, 0, 1]:
    print(sign(x))

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 10/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

negative
zero
positive

We will often define functions to take optional keyword arguments, like this:

def hello(name, loud=False):
    if loud:
        print('HELLO, {}'.format(name.upper()))
    else:
        print('Hello, {}!'.format(name))

hello('Bob')
hello('Fred', loud=True)

Hello, Bob!
HELLO, FRED

Classes

The syntax for defining classes in Python is straightforward:

class Greeter:

    # Constructor
    def __init__(self, name):
        self.name = name  # Create an instance variable

    # Instance method
    def greet(self, loud=False):
        if loud:
          print('HELLO, {}'.format(self.name.upper()))
        else:
          print('Hello, {}!'.format(self.name))

g = Greeter('Fred')  # Construct an instance of the Greeter class
g.greet()            # Call an instance method; prints "Hello, Fred"
g.greet(loud=True)   # Call an instance method; prints "HELLO, FRED!"

Hello, Fred!
HELLO, FRED

Numpy

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 11/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

Numpy is the core library for scientific computing in Python. It provides a high-performance
multidimensional array object, and tools for working with these arrays. If you are already familiar
with MATLAB, you might find this tutorial useful to get started with Numpy.

To use Numpy, we first need to import the numpy package:

import numpy as np

Arrays

A numpy array is a grid of values, all of the same type, and is indexed by a tuple of nonnegative
integers. The number of dimensions is the rank of the array; the shape of an array is a tuple of
integers giving the size of the array along each dimension.

We can initialize numpy arrays from nested Python lists, and access elements using square
brackets:

a = np.array([1, 2, 3])  # Create a rank 1 array
print(type(a), a.shape, a[0], a[1], a[2])
a[0] = 5                 # Change an element of the array
print(a)                  

<class 'numpy.ndarray'> (3,) 1 2 3

[5 2 3]

b = np.array([[1,2,3],[4,5,6]])   # Create a rank 2 array
print(b)

[[1 2 3]
[4 5 6]]

print(b.shape)
print(b[0, 0], b[0, 1], b[1, 0])

(2, 3)
1 2 4

Numpy also provides many functions to create arrays:

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 12/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

a = np.zeros((2,2))  # Create an array of all zeros
print(a)

[[0. 0.]
[0. 0.]]

b = np.ones((1,2))   # Create an array of all ones
print(b)

[[1. 1.]]

c = np.full((2,2), 7) # Create a constant array
print(c)

[[7 7]
[7 7]]

d = np.eye(2)        # Create a 2x2 identity matrix
print(d)

[[1. 0.]
[0. 1.]]

e = np.random.random((2,2)) # Create an array filled with random values
print(e)

[[0.8690054 0.57244319]
[0.29647245 0.81464494]]

Array indexing

Numpy offers several ways to index into arrays.

Slicing: Similar to Python lists, numpy arrays can be sliced. Since arrays may be multidimensional,
you must specify a slice for each dimension of the array:

import numpy as np

# Create the following rank 2 array with shape (3, 4)
# [[ 1  2  3  4]
#  [ 5  6  7  8]
#  [ 9 10 11 12]]
a = np.array([[1,2,3,4], [5,6,7,8], [9,10,11,12]])

# Use slicing to pull out the subarray consisting of the first 2 rows
https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 13/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

# and columns 1 and 2; b is the following array of shape (2, 2):
# [[2 3]
#  [6 7]]
b = a[:2, 1:3]
print(b)
[[2 3]
[6 7]]

A slice of an array is a view into the same data, so modifying it will modify the original array.

print(a[0, 1])
b[0, 0] = 77    # b[0, 0] is the same piece of data as a[0, 1]
print(a[0, 1]) 

2
77

You can also mix integer indexing with slice indexing. However, doing so will yield an array of lower
rank than the original array. Note that this is quite different from the way that MATLAB handles array
slicing:

# Create the following rank 2 array with shape (3, 4)
a = np.array([[1,2,3,4], [5,6,7,8], [9,10,11,12]])
print(a)

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

Two ways of accessing the data in the middle row of the array. Mixing integer indexing with slices
yields an array of lower rank, while using only slices yields an array of the same rank as the original
array:

row_r1 = a[1, :]    # Rank 1 view of the second row of a  
row_r2 = a[1:2, :]  # Rank 2 view of the second row of a
row_r3 = a[[1], :]  # Rank 2 view of the second row of a
print(row_r1, row_r1.shape)
print(row_r2, row_r2.shape)
print(row_r3, row_r3.shape)

[5 6 7 8] (4,)
[[5 6 7 8]] (1, 4)
[[5 6 7 8]] (1, 4)

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 14/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

# We can make the same distinction when accessing columns of an array:
col_r1 = a[:, 1]
col_r2 = a[:, 1:2]
print(col_r1, col_r1.shape)
print()
print(col_r2, col_r2.shape)

[ 2 6 10] (3,)

[[ 2]
[ 6]
[10]] (3, 1)

Integer array indexing: When you index into numpy arrays using slicing, the resulting array view will
always be a subarray of the original array. In contrast, integer array indexing allows you to construct
arbitrary arrays using the data from another array. Here is an example:

a = np.array([[1,2], [3, 4], [5, 6]])

# An example of integer array indexing.
# The returned array will have shape (3,) and 
print(a[[0, 1, 2], [0, 1, 0]])

# The above example of integer array indexing is equivalent to this:
print(np.array([a[0, 0], a[1, 1], a[2, 0]]))

[1 4 5]
[1 4 5]

# When using integer array indexing, you can reuse the same
# element from the source array:
print(a[[0, 0], [1, 1]])

# Equivalent to the previous integer array indexing example
print(np.array([a[0, 1], a[0, 1]]))

[2 2]
[2 2]

One useful trick with integer array indexing is selecting or mutating one element from each row of a
matrix:

# Create a new array from which we will select elements
a = np.array([[1,2,3], [4,5,6], [7,8,9], [10, 11, 12]])
print(a)

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 15/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

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

# Create an array of indices
b = np.array([0, 2, 0, 1])

# Select one element from each row of a using the indices in b
print(a[np.arange(4), b])  # Prints "[ 1  6  7 11]"

[ 1 6 7 11]

# Mutate one element from each row of a using the indices in b
a[np.arange(4), b] += 10
print(a)

[[11 2 3]
[ 4 5 16]
[17 8 9]
[10 21 12]]

Boolean array indexing: Boolean array indexing lets you pick out arbitrary elements of an array.
Frequently this type of indexing is used to select the elements of an array that satisfy some
condition. Here is an example:

import numpy as np

a = np.array([[1,2], [3, 4], [5, 6]])

bool_idx = (a > 2)  # Find the elements of a that are bigger than 2;
                    # this returns a numpy array of Booleans of the same
                    # shape as a, where each slot of bool_idx tells
                    # whether that element of a is > 2.

print(bool_idx)

[[False False]
[ True True]
[ True True]]

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 16/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

# We use boolean array indexing to construct a rank 1 array
# consisting of the elements of a corresponding to the True values
# of bool_idx
print(a[bool_idx])

# We can do all of the above in a single concise statement:
print(a[a > 2])

[3 4 5 6]
[3 4 5 6]

For brevity we have left out a lot of details about numpy array indexing; if you want to know more
you should read the documentation.

Datatypes

Every numpy array is a grid of elements of the same type. Numpy provides a large set of numeric
datatypes that you can use to construct arrays. Numpy tries to guess a datatype when you create
an array, but functions that construct arrays usually also include an optional argument to explicitly
specify the datatype. Here is an example:

x = np.array([1, 2])  # Let numpy choose the datatype
y = np.array([1.0, 2.0])  # Let numpy choose the datatype
z = np.array([1, 2], dtype=np.int64)  # Force a particular datatype

print(x.dtype, y.dtype, z.dtype)

int64 float64 int64

You can read all about numpy datatypes in the documentation.

Array math

Basic mathematical functions operate elementwise on arrays, and are available both as operator
overloads and as functions in the numpy module:

x = np.array([[1,2],[3,4]], dtype=np.float64)
y = np.array([[5,6],[7,8]], dtype=np.float64)

# Elementwise sum; both produce the array

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 17/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

print(x + y)
[[ 6. 8.]
print(np.add(x, y))
[10. 12.]]
[[ 6. 8.]
[10. 12.]]

# Elementwise difference; both produce the array
print(x - y)
print(np.subtract(x, y))

[[-4. -4.]
[-4. -4.]]
[[-4. -4.]
[-4. -4.]]

# Elementwise product; both produce the array
print(x * y)
print(np.multiply(x, y))

[[ 5. 12.]
[21. 32.]]
[[ 5. 12.]
[21. 32.]]

# Elementwise division; both produce the array
# [[ 0.2         0.33333333]
#  [ 0.42857143  0.5       ]]
print(x / y)
print(np.divide(x, y))

[[0.2 0.33333333]
[0.42857143 0.5 ]]
[[0.2 0.33333333]
[0.42857143 0.5 ]]

# Elementwise square root; produces the array
# [[ 1.          1.41421356]
#  [ 1.73205081  2.        ]]
print(np.sqrt(x))

[[1. 1.41421356]
[1.73205081 2. ]]

Note that unlike MATLAB, * is elementwise multiplication, not matrix multiplication. We instead
use the dot function to compute inner products of vectors, to multiply a vector by a matrix, and to
multiply matrices. dot is available both as a function in the numpy module and as an instance
method of array objects:

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 18/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

x = np.array([[1,2],[3,4]])
y = np.array([[5,6],[7,8]])

v = np.array([9,10])
w = np.array([11, 12])

# Inner product of vectors; both produce 219
print(v.dot(w))
print(np.dot(v, w))

219
219

You can also use the @ operator which is equivalent to numpy's dot operator.

print(v @ w)

219

# Matrix / vector product; both produce the rank 1 array [29 67]
print(x.dot(v))
print(np.dot(x, v))
print(x @ v)

[29 67]
[29 67]
[29 67]

# Matrix / matrix product; both produce the rank 2 array
# [[19 22]
#  [43 50]]
print(x.dot(y))
print(np.dot(x, y))
print(x @ y)

[[19 22]
[43 50]]
[[19 22]
[43 50]]
[[19 22]
[43 50]]

Numpy provides many useful functions for performing computations on arrays; one of the most
useful is sum :

x = np.array([[1,2],[3,4]])

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 19/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

print(np.sum(x))  # Compute sum of all elements; prints "10"
print(np.sum(x, axis=0))  # Compute sum of each column; prints "[4 6]"
print(np.sum(x, axis=1))  # Compute sum of each row; prints "[3 7]"
10
[4 6]
[3 7]

You can find the full list of mathematical functions provided by numpy in the documentation.

Apart from computing mathematical functions using arrays, we frequently need to reshape or
otherwise manipulate data in arrays. The simplest example of this type of operation is transposing
a matrix; to transpose a matrix, simply use the T attribute of an array object:

print(x)
print("transpose\n", x.T)

[[1 2]
[3 4]]
transpose
[[1 3]
[2 4]]

v = np.array([[1,2,3]])
print(v )
print("transpose\n", v.T)

[[1 2 3]]
transpose
[[1]
[2]
[3]]

Broadcasting

Broadcasting is a powerful mechanism that allows numpy to work with arrays of different shapes
when performing arithmetic operations. Frequently we have a smaller array and a larger array, and
we want to use the smaller array multiple times to perform some operation on the larger array.

For example, suppose that we want to add a constant vector to each row of a matrix. We could do it
like this:

# We will add the vector v to each row of the matrix x,
# storing the result in the matrix y
x = np.array([[1,2,3], [4,5,6], [7,8,9], [10, 11, 12]])
v = np.array([1, 0, 1])

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 20/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

y = np.empty_like(x)   # Create an empty matrix with the same shape as x

# Add the vector v to each row of the matrix x with an explicit loop
for i in range(4):
    y[i, :] = x[i, :] + v

print(y)
[[ 2 2 4]
[ 5 5 7]
[ 8 8 10]
[11 11 13]]

This works; however when the matrix x is very large, computing an explicit loop in Python could be
slow. Note that adding the vector v to each row of the matrix x is equivalent to forming a matrix vv
by stacking multiple copies of v vertically, then performing elementwise summation of x and vv .
We could implement this approach like this:

vv = np.tile(v, (4, 1))  # Stack 4 copies of v on top of each other
print(vv)                # Prints "[[1 0 1]
                         #          [1 0 1]
                         #          [1 0 1]
                         #          [1 0 1]]"

[[1 0 1]
[1 0 1]
[1 0 1]
[1 0 1]]

y = x + vv  # Add x and vv elementwise
print(y)

[[ 2 2 4]
[ 5 5 7]
[ 8 8 10]
[11 11 13]]

Numpy broadcasting allows us to perform this computation without actually creating multiple
copies of v. Consider this version, using broadcasting:

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 21/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

import numpy as np

# We will add the vector v to each row of the matrix x,
# storing the result in the matrix y
x = np.array([[1,2,3], [4,5,6], [7,8,9], [10, 11, 12]])
v = np.array([1, 0, 1])
y = x + v  # Add v to each row of x using broadcasting
print(y)

[[ 2 2 4]
[ 5 5 7]
[ 8 8 10]
[11 11 13]]

The line y = x + v works even though x has shape (4, 3) and v has shape (3,) due to
broadcasting; this line works as if v actually had shape (4, 3) , where each row was a copy of v ,
and the sum was performed elementwise.

Broadcasting two arrays together follows these rules:

1. If the arrays do not have the same rank, prepend the shape of the lower rank array with 1s until
both shapes have the same length.
2. The two arrays are said to be compatible in a dimension if they have the same size in the
dimension, or if one of the arrays has size 1 in that dimension.
3. The arrays can be broadcast together if they are compatible in all dimensions.
4. After broadcasting, each array behaves as if it had shape equal to the elementwise maximum
of shapes of the two input arrays.
5. In any dimension where one array had size 1 and the other array had size greater than 1, the
first array behaves as if it were copied along that dimension

If this explanation does not make sense, try reading the explanation from the documentation or this
explanation.

Functions that support broadcasting are known as universal functions. You can find the list of all
universal functions in the documentation.

Here are some applications of broadcasting:

# Compute outer product of vectors
v = np.array([1,2,3])  # v has shape (3,)
w = np.array([4,5])    # w has shape (2,)
# To compute an outer product, we first reshape v to be a column
# vector of shape (3, 1); we can then broadcast it against w to yield
# an output of shape (3, 2), which is the outer product of v and w:

print(np.reshape(v, (3, 1)) * w)

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 22/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

[[ 4 5]
[ 8 10]
[12 15]]

# Add a vector to each row of a matrix
x = np.array([[1,2,3], [4,5,6]])
# x has shape (2, 3) and v has shape (3,) so they broadcast to (2, 3),
# giving the following matrix:

print(x + v)

[[2 4 6]
[5 7 9]]

# Add a vector to each column of a matrix
# x has shape (2, 3) and w has shape (2,).
# If we transpose x then it has shape (3, 2) and can be broadcast
# against w to yield a result of shape (3, 2); transposing this result
# yields the final result of shape (2, 3) which is the matrix x with
# the vector w added to each column. Gives the following matrix:

print((x.T + w).T)

[[ 5 6 7]
[ 9 10 11]]

# Another solution is to reshape w to be a row vector of shape (2, 1);
# we can then broadcast it directly against x to produce the same
# output.
print(x + np.reshape(w, (2, 1)))

[[ 5 6 7]
[ 9 10 11]]

# Multiply a matrix by a constant:
# x has shape (2, 3). Numpy treats scalars as arrays of shape ();
# these can be broadcast together to shape (2, 3), producing the
# following array:
print(x * 2)

[[ 2 4 6]
[ 8 10 12]]

Broadcasting typically makes your code more concise and faster, so you should strive to use it
where possible.

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 23/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

This brief overview has touched on many of the important things that you need to know about
numpy, but is far from complete. Check out the numpy reference to find out much more about
numpy.

Matplotlib

Matplotlib is a plotting library. In this section give a brief introduction to the matplotlib.pyplot
module, which provides a plotting system similar to that of MATLAB.

import matplotlib.pyplot as plt

By running this special iPython command, we will be displaying plots inline:

%matplotlib inline

Plotting

The most important function in matplotlib is plot, which allows you to plot 2D data. Here is a
simple example:

# Compute the x and y coordinates for points on a sine curve
x = np.arange(0, 3 * np.pi, 0.1)
y = np.sin(x)

# Plot the points using matplotlib
plt.plot(x, y)

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 24/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

[<matplotlib.lines.Line2D at 0x7f0f3a0b4208>]
With just a little bit of extra work we can easily plot multiple lines at once, and add a title, legend,
and axis labels:

y_sin = np.sin(x)
y_cos = np.cos(x)

# Plot the points using matplotlib
plt.plot(x, y_sin)
plt.plot(x, y_cos)
plt.xlabel('x axis label')
plt.ylabel('y axis label')
plt.title('Sine and Cosine')
plt.legend(['Sine', 'Cosine'])

<matplotlib.legend.Legend at 0x7f0f39c04780>

Subplots

You can plot different things in the same figure using the subplot function. Here is an example:

# Compute the x and y coordinates for points on sine and cosine curves
x = np.arange(0, 3 * np.pi, 0.1)
y_sin = np.sin(x)
y_cos = np.cos(x)

# Set up a subplot grid that has height 2 and width 1,
# and set the first such subplot as active.
plt.subplot(2, 1, 1)

# Make the first plot
plt.plot(x, y_sin)
https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 25/26
2/24/23, 12:00 AM Copy of colab-tutorial.ipynb - Colaboratory

plt.title('Sine')

# Set the second subplot as active, and make the second plot.
plt.subplot(2, 1, 2)
plt.plot(x, y_cos)
plt.title('Cosine')

# Show the figure.
plt.show()

https://colab.research.google.com/drive/1Z5ItdLyrRIUR8UMHcP9EYbHN56H_2OfO#scrollTo=0vJLt3JRL9eR 26/26