Useful Functions

It's about time that we wrote some Python! Everything from here on out is about tricks in Python that will help you in 6.101.

String Manipulation

Strings in Python have a super useful library of methods that you can call on them, and this section will explain some of them. For the full list, see this link

split, join, replace

split, join, and replace are three string functions you ought to know.

split is simple. It splits a string into a list of strings by a certain character. By default, it will just split on spaces, though you can specify a character to split on.

>>> "1 2 3".split()
['1', '2', '3']

>>> "1,2,3".split(",")
['1', '2', '3']

join does the exact opposite. It inserts a string in between strings in an iterable¹. This is helpful when you have a list of strings (or list comprehension) that you want to merge together into a string. You can use the empty string (""), if you want to join the elements together without anything in between.

>>> ",".join(["a", "b", "c"])
'a,b,c'

>>> "".join(['x', 'y', 'z'])
'xyz'

>>> " -> ".join(str(i) for i in range(10))
'0 -> 1 -> 2 -> 3 -> 4 -> 5 -> 6 -> 7 -> 8 -> 9'

replace is also super simple. If it finds an occurrence of a string, it replaces it with the new value and returns the modified string. If it doesn't find it, it returns the original string.

>>> "This guide is for 6.100 students.".replace("100", "101")
'This guide is for 6.101 students.'

>>> "Hello!".replace("Hi", "Howdy")
'Hello!'

f-strings

F-strings, or Formatted String Literals, in Python let you embed the value of variables in strings in your program. This concept is called string interpolation.

The most basic use of f-strings requires putting an f before the start of your Python string, and then, you can embed variables in your string using curly braces ({}).

>>> name = "Cameron"

>>> f"Hello, {name}!"
'Hello, Cameron!'

Putting an equals sign after the variable name prints out the name of the variable as well, and this can be extremely useful when you're debugging.

>>> f"Variables: {name=}"
"Variables: name='Cameron'"

List Utilities

Python has a ton of super useful built-in functions that you should know about. A comprehensive list can be found here, but we'll be going into detail about some of them below.

Many of these utilities reference Python iterables. An iterable just means anything you can iterate over, and this includes lists, sets, tuples, dicts, generators, and any class that implements the __iter__ method. These methods should help save you from writing tons of for loops.

The + operator and list.extend

The + operator on lists is a convenient way to combine two lists together, returning a new list with all elements from both lists.

>>> [1, 2] + [3, 4]
[1, 2, 3, 4]

If we want to update a list in place, we can use list.extend, which is faster than using the + operator. list.extend is more flexible in that it lets us specify any iterable instead of only lists.

>>> a = [1, 2]
>>> a.extend([3, 4])
>>> a.extend({5, 6})  # set
>>> a.extend((7, 8))  # tuple
>>> a
[1, 2, 3, 4, 5, 6, 7, 8]

The += operator is actually identical to list.extend, not the + operator. This means that a = a + [3, 4] and a += [3, 4] behave differently! The += operator extends the underlying list, while the + operator creates a new, combined list and assigns that to a. Here's an example of this behavior in action.

>>> a = [1, 2]
>>> b = [1, 2]
>>> a2, b2 = a, b  # Alias a2 and b2 to a and b, respectively
>>> a += [3, 4]
>>> b = b + [3, 4]
>>> a2
[1, 2, 3, 4]
>>> b2
[1, 2]

Notice that the += operator extended the list, so this change was also reflected in a2. Meanwhile, the + operator made a new list and assigned it to b, so b2 was left pointing to the original list.

min, max

min finds the minimum of multiple elements, and max finds the maximum. They take anything that can be compared, like ints and floats, or you can pass them strings to use alphabetical order. They can be used in two different ways. You can pass in multiple arguments, like so:

>>> max(7, 5, 8, 5.5)
8

>>> min("banana", "cat", "apple")
'apple'

You can also give an iterable of items, like a list, set, tuple, or comprehension, to find the minimum or maximum of as a single argument.

>>> max([7, 5, 8, 5,5])
8

>>> min(("banana", "cat", "apple"))
'apple'

sorted, reversed

sorted returns a copy of an iterable (like a list or tuple) as a sorted list, while list.sort and list.reverse work in-place.

>>> sorted(["python", "typescript", "html"])
['html', 'python', 'typescript']

reversed returns an iterable, so you need to call list or tuple on it in order to convert it to something you can use.

>>> reversed(("first", "second", "third"))
<reversed object at 0x10099b490>
>>> list(reversed(("first", "second", "third")))
['third', 'second', 'first']

The key argument

If you have some data that can't be compared directly, min and max (as well as sorted and reversed) both take a key= argument, which allows you to specify a lambda function that returns a value that will be compared instead. This works especially well for tuples of data.

>>> max(("foo", 4), ("bar", 7), ("baz", 2), key=lambda v: v[1])
('bar', 7)

>>> min("longer string", "long string", "string", "str", key=len)
'str'

We can also call sorted with the key argument.

>>> sorted(["longer string", "long string", "string", "str"], key=len)
['str', 'string', 'long string', 'longer string']

If you want to sort in place, list.sort also accepts the same key argument.

>>> data = [("foo", 4), ("bar", 7), ("baz", 2)]
>>> data.sort(key=lambda v: v[1])
>>> data
[('baz', 2), ('foo', 4), ('bar', 7)]

Sorting (or using min or max) tuples without a key parameter sorts based on the first element in the tuples.

>>> data.sort()
>>> data
[('bar', 7), ('baz', 2), ('foo', 4)]

any, all

any and all are functions that check if any or all values in an iterable are truthy.² Here's some examples:

>>> any([True, False, False])
True
>>> True or False or False  # The same as combining the elements with or
True

>>> all([True, False, False])
False
>>> True and False and False  # The same as combining the elements with and
False

These can be especially helpful with comprehensions, or types that aren't booleans.

>>> all(len(s) > 3 for s in ["apple", "banana", "kiwi"])
True
>>> any([0, 2, 0])
True

sum

sum takes in an iterable and sums up its components. It takes comprehensions as well, so it comes in handy often. Try to use it instead of a for loop when you can.

>>> sum([1,2,3])
6
>>> sum(i for i in range(10))  # Sum of 1 to 10
45

zip, enumerate

zip takes in any number of iterables and combines them together into a single iterable of tuples. You need to convert it to a list or a tuple if you're not going to iterate over it. This is helpful if you have two or more lists that correspond to each other and want to iterate over them at the same time.

>>> labels = ['apples', 'bananas', 'kiwis']
>>> inventory = [12, 30, 5]
>>> zip(labels, inventory)
<zip object at 0x100c94f80>
>>> list(zip(labels, inventory))
[('apples', 12), ('bananas', 30), ('kiwis', 5)]

What's cool is that this is the same format as dict.items, so the dict constructor takes in key-value pairs in this format. So, it's super easy to make a dictionary from our zip output.

>>> dict(zip(labels, inventory))
{'apples': 12, 'bananas': 30, 'kiwis': 5}

Have you ever been iterating over a list and wanted its indices as well as its values? enumerate is just the function you need. It takes in an iterable and returns an iterable of tuple pairs (index, value).

>>> list(enumerate(labels))
[(0, 'apples'), (1, 'bananas'), (2, 'kiwis')]
>>> for index, label in enumerate(labels):
...     print(f"{index}: {label}")
0: apples
1: bananas
2: kiwis

Tuple unpacking

PLEASE USE THIS, THANKS.

❤️ LOVE, YOUR 6.101 LAs

Tuple unpacking is an incredibly useful feature that lets you break apart tuples into their elements and assign each element in a tuple to a separate variable. If you have a tuple, you can separate variables by commas to assign each variable to that element in a tuple, like so. This behavior is referred to as "unpacking" the tuple.

>>> number, letter = (1, "A")
>>> number
1
>>> letter
'A'

This feature gets useful once we use it to work with data that's structured in tuples. Without tuple unpacking, you might be inclined to access the flight information in this example by indexing into the tuple, like this:

>>> flights = [(1234, "BOS", "LAX"), (5678, "JFK", "ATL")]
>>> for flight in flights:
...     print(f"Flight {flight[0]} departs {flight[1]} for {flight[2]}")
Flight 1234 departs BOS for LAX
Flight 5678 departs JFK for ATL

However, a much more readable option would be to break apart, or "unpack", the flight tuple in the for loop, like this:

>>> for flight in flights:
...     num, origin, dest = flight
...     print(f"Flight {num} departs {origin} for {dest}")
Flight 1234 departs BOS for LAX
Flight 5678 departs JFK for ATL

Now, num is always assigned to the flight number, origin is always assigned to the flight's origin, and dest is always assigned to the flight destination. This makes the code much more readable for when anyone is debugging your code.

Python also has a helpful shorthand for this, which moves the unpacking into the first line of the for loop, like this:

>>> for num, origin, dest in flights:
...     print(f"Flight {num} departs {origin} for {dest}")
Flight 1234 departs BOS for LAX
Flight 5678 departs JFK for ATL

dict and set Utilities

As you'll learn in 6.101, dict and set in Python have special properties that make them useful. For a dictionary, we can map arbitrary keys to arbitrary values, so that it's extremely fast to look up and retrieve a value. In sets, we can check inclusion extremely quickly to see if a given value is in a set.

In this section, you'll learn some handy tricks for working with dict and set.

frozenset

Unfortunately, we can only put immutable objects inside sets. Because sets are mutable, we can't put sets inside sets, which is something you might want to do in a 6.101 lab or two. frozenset fixes that problem for us because frozensets are immutable, so we can put them inside sets.

You can create a frozenset with its constructor from any iterable, like another set, list, generator, or even a string.

>>> alphabet = frozenset("abcdefghijklmnopqrstuvwxyz")
>>> alphabet
frozenset({'q', 'm', 'w', 'x', 'h', 's', 't', 'a', 'n', 'j', 'c', 'f', 'g', 'u', 'd', 'l', 'k', 'p', 'r', 'o', 'e', 'b', 'z', 'v', 'i', 'y'})

Say I have a class of students that I arrange into groups of two. The specifics of how I make the groups don't matter all that much, just that we have a list of lists at this point.

>>> students = ["John", "Jack", "Sally", "Susan", "Alice", "Bob"]
>>> random.shuffle(students)
>>> groups = [students[i:i+2] for i in range(0, len(students), 2)]
>>> groups
[['John', 'Susan'], ['Bob', 'Sally'], ['Alice', 'Jack']]

If I want to find out whether John and Susan are paired up together, I need to do a couple things:

1. First, I have to loop over all the groups in the list, which could be very slow if there are a lot of groups.
2. Next, for each group, I need to check if both John and Susan are in it, considering that they could be in any order.

A better approach would be to convert all the inner lists to sets, like this:

>>> groups = [set(g) for g in groups]
>>> groups
[{'John', 'Susan'}, {'Bob', 'Sally'}, {'Alice', 'Jack'}]
>>> {'Susan', 'John'} in groups
True
>>> {'John', 'Susan'} in groups
True

You'll notice that we can use this more elegant syntax with the in operator, and it doesn't matter what order "John" and "Susan" are when we check. This is more convenient, but the in operator on lists loops over the list to check for inclusion, which could be slow if the list is very long.

The best approach is to use frozensets inside sets. Fortunately, this is super easy to adopt.

>>> groups = {frozenset(g) for g in groups}
>>> groups
{frozenset({'Alice', 'Jack'}), frozenset({'John', 'Susan'}), frozenset({'Bob', 'Sally'})}
>>> {'John', 'Susan'} in groups
True
>>> {'Susan', 'John'} in groups
True

This is way faster because each in check takes a constant amount of time, regardless of how full the set is. Notice also that we can still use regular sets to check inside our list of frozensets. This is because Python treats frozensets and sets the same, and you can even check if their contents are equal:

>>> {'a'} == frozenset({'a'})
True

dict.get

get is a dict method we can use to retrieve a value from a dictionary. But you might be wondering, can't we already do that with d[key]? The difference is that dict.get doesn't raise a KeyError when key isn't in the dictionary (it returns None instead), and we can even specify a default value to get returned.

>>> inventory = {'carrots': 3, 'cucumbers': 7}
>>> inventory['peppers']  # booo, KeyError :(
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
KeyError: 'peppers'
>>> inventory.get('peppers')  # No KeyError, just None!
>>> inventory.get('peppers', 0)  # No KeyError, now we get 0!
0

dict.setdefault

setdefault comes in handy when we have dictionaries with mutable values that we want to either initialize or change. Say I have a list of flights, and I'm trying to make a dictionary mapping origins to the destinations:

>>> flights = [('BOS', 'JFK'), ('BOS', 'ATL'), ('JFK', 'LAX'), ('LAX', 'SFO'), ('BOS', 'LAX'), ('JFK', 'DFW')]
>>> destinations = {}
>>> for origin, dest in flights:
...     if origin in destinations:
...             destinations[origin].add(dest)
...     else:
...             destinations[origin] = {dest}
>>> destinations
{'BOS': {'ATL', 'JFK', 'LAX'}, 'JFK': {'DFW', 'LAX'}, 'LAX': {'SFO'}}

This loop is made way more complicated by having to check whether we need to create a new set in the dictionary for a given key. dict.setdefault lets us combine these two steps into one by creating a new set for us automatically if one doesn't exist already.

>>> for origin, dest in flights:
...     destinations.setdefault(origin, set()).add(dest)
>>> destinations
{'BOS': {'ATL', 'JFK', 'LAX'}, 'JFK': {'DFW', 'LAX'}, 'LAX': {'SFO'}}

The | operator, dict.update, and set.update

The | operator is the "intersection" operator for sets and dictionaries in Python. In short, it lets you merge together two sets, or two dictionaries. In the case of a dict, the values in the right dict take precedence.

>>> {'a'} | {'b', 'c'}
{'b', 'a', 'c'}
>>> {'a': 1} | {'b': 2, 'a': 3}
{'a': 3, 'b': 2}

Like with other operators in Python, we can combine the intersection operator with the = assignment operator, using the |= operator to "update" a dictionary or set with new values.

>>> s = {1,2,3}
>>> s |= {4,5,6}
>>> s
{1, 2, 3, 4, 5, 6}
>>> d = {'a': 1}
>>> d |= {'b': 2}
>>> d
{'a': 1, 'b': 2}

The dict.update and set.update methods are identical to using the |= operator.

>>> s.update({7,8,9})
>>> s
{1, 2, 3, 4, 5, 6, 7, 8, 9}
>>> d.update({'c':3})
>>> d
{'a': 1, 'b': 2, 'c': 3}

dict.update also overwrites values in the dictionary with the new values.

>>> d.update({'c': 30, 'd': 40})
>>> d
{'a': 1, 'b': 2, 'c': 30, 'd': 40}

dict.items

dict.items is a method that yields elements in a dictionary as key-value pairs, and you should use it if you need to access both the keys and values of a dictionary in a for loop. Here's how it works:

>>> inventory = {'carrots': 3, 'cucumbers': 7, 'broccoli': 10}
>>> for key, value in inventory.items():
...     print(f"There are {value} pieces of {key} in the store.")
There are 3 pieces of carrots in the store.
There are 7 pieces of cucumbers in the store.
There are 10 pieces of broccoli in the store.

---

1. An iterable is just anything that can be iterated on. It includes lists, tuples, generators, or classes that implement the __iter__ method. ↩

2. If you need a refresher on what a truthy value in Python means, check out this tutorial. ↩