Once you’ve mastered loops and conditionals, we can talk about comprehensions. Comprehensions are clever short-hand statements used to create sequence data structures. for instance, we can fill a list by manipulating data from another list quickly using list comprehension:
a = [1, 2, 3]
b = [i**2 for i in a]
c = [i**2 for i in a if i % 2 == 0]
We can see that the syntax for a list comprehension uses the list literal
operators [ and ] to enclose a deconstructed for loop:
# Unconditional list comprehension
[loop_body for iterator_variable in sequence]
# An optional conditional can be included to modify
# when a new item is added to the new list.
[loop_body for iterator_variable in sequence if conditional]
Tuples can be formed the same way using tuple comprehension. The only syntax difference from list comprehensions is the use of the tuple literal operators:
b = (i**2 for i in range(10))
c = (i**2 for i in range(10) if i % 2 == 0)
Likewise, set comprehensions are formed using the same syntax but with the
set literal operators { and }:
superheroes = ['spider man', 'wolverine', 'professor x',
'batman', 'jean grey', 'catwoman',
'superman', 'green lantern']
one_name_heroes = {name for name in superheroes if ' ' not in name}
Dictionaries also have a comprehension but it is slightly more complicated than the others. It looks like this:
{key : value for item in sequence}
For instance,
# Make a dictionary from a string:
simple_dict = {i : j for i, j in enumerate('abcd')}
# To swap keys and values:
swapped = {j : i for i, j in simple_dict.items()}
Well, comprehensions are great. The real question is when not to use them? You should avoid using a comprehension when the whole expression is much longer than a single 80-character long line. When you try to pack so much code into a single line, it makes the program hard to follow. Otherwise, use them whenever you can.