List Comprehensions

In data work, you're constantly building new lists. Say you have a list of songs:

tracks = [
    {"track_id": "T-100", "plays": 2},
    {"track_id": "T-101", "plays": 5},
    {"track_id": "T-102", "plays": 4},
]

And you want a new list with just the number of "plays" for each track. You can do that with a normal for loop:

play_counts = []
for track in tracks:
    play_counts.append(track["plays"])
print(play_counts)
# [2, 5, 4]

It works, but it's a lot of boilerplate. Python gives us a shorter way with list comprehensions.

play_counts = [track["plays"] for track in tracks]
print(play_counts)
# [2, 5, 4]

It says:

• take track["plays"]
• for each track
• in the tracks list
• and give me a new list with those values (play_counts)

The syntax is:

[expression for item in iterable]

And if you want to filter as well:

[expression for item in iterable if condition]

For example:

long_play_counts = [track["plays"] for track in tracks if track["plays"] > 3]
print(long_play_counts)
# [5, 4]

Not only are comprehensions more concise, but they're also often more performant than a normal loop because the Python interpreter can optimize them more easily under the hood.