I’m learning a brand new programming language. That should probably say, new for me.Because it’s Python. I haven’t touched it before. Yes yes - I’m slow, old and everything. But I am still learning new things… so I’m not all in all bad.
Ok - Python is for the most part very nice to work with and reading and writing code has been easy so far. Until I saw a backwards for-loop. It just looked… weird.
Turns out it was not a for-loop at all, but rather a list comprehension.
My need was not super basic, and it bothered me that I couldn’t solve it nicer than I did, so I wanted see if I can understand how I could make it better. Hence this blog post.
List comprehensions
Let’s first just checkout the most basic case. List comprehensions are pretty cool. Here’s the basic structure:
newList = [ expression(element) for element in oldList if condition ]
It’s basically a way to create list from list. This can be used for transformations, filtering, mapping and all kinds of coolness. Like some of the array methods in JavaScript, or LINQ in C# etc.
Let’s see one in action. :
someNumbers = [1, 3, 56, 2, 4]
theNumbersDoubled = [num * 2 for num in someNumbers if num < 50]
print(theNumbersDoubled) # [2, 6, 4, 8]
Pretty cool, huh?
First let me just say that I like that you wrap the whole thing in a []. It’s like saying: here’s the recipe for the new list I’m want to make. Inside is the declarative code.
The thing that threw me off is the backwards (I think…) structure. The for num in someNumbers comes at the end. It can, optionally have an if-statement as you see me do there if num < 50.
The projection, restructuring of the code, comes first; num * 2. num is what I call the variable I’m iterating over.
Let’s make something a bit more advanced, by having object in the array, so that we can show a bit more of the projection powers. Also, I will learn a bit more Python.
class person:
def __init__(self, name, birthYear):
self.name = name
self.birthYear = birthYear
theHammarbergs = [
person('Marcus', 1972),
person('Elin', 1977),
person('Albert', 2008),
person('Arvid', 2010),
person('Gustav', 2020)
]
current_year = date.today().year
adults = [p.name for p in theHammarbergs if current_year - p.birthYear >= 18]
print(adults)
First - I created a class for person and then filled up an array with the people in my family. Yes - we have twins.
(I then found out how to get the current year… Yes I’m new at this)
Finally I could now write a little expression that filters the array and selects only the name from those people. Beautiful.
Making dictionaries
My problem included using a Dictionary. Or rather, I needed to output a Dictionary with a key and value for a number of items… Yeah it’s a bit complicated, but like this: I have one list of strings that are the keys, and then a list of list of strings that are the rows. From this I need to create a list of Dictionaries.
First things first - here’s a list of strings, in Python:
columns: list[str] = ['Name', 'Age', 'Shoe-size']
Then a list of list of strings, for the values:
rows: list[list[str]] = [
['Marcus', '1972', '46'],
['Elin', '1977', '37'],
['Albert', '2008', '40'],
['Arvid', '2010', '38'],
['Gustav', '2010', '38']
]
And my goal is to create a list of dictionary of string keys and string values (let’s make it easy). It will look something like this:
result: list[dict[str, str]] = [
{"Name": "Marcus", "BirthYear": "1972", "Shoe-size": "46"},
{"Name": "Elin", "BirthYear": "1977", "Shoe-size": "38"},
# and so on
]
Now, there’s a very cool way of creating a Dictionary from two arrays, by using zip like this:
result = dict(zip(["a", "b", "c"], [1, 2, 3]))
print(result) # {'a': 1, 'b': 2, 'c': 3}
We could use this and create a dictionary for every row. That makes the whole code like this:
results = [dict(zip(columns, row)) for row in rows]
print(results)
Summary
Using List comprehensions in Python makes your code much smaller and more effective. But it can get some getting used to read them. I think we will be friends.