blog/content/code/you-dont-need-range.md

+++ title = "You Don't Need range()" date = 2020-04-16

[taxonomies] tags = ["code", "python", "range"] +++

Beginners in Python tend to use range() for iterating over lists. This is not really necessary.

When people start programming Python, they tend to use constructions coming from other languages, so they iterate over a list with something like:

a_list = [1, 2, 3, 4]
for i in range(len(a_list)):
    print(a_list[i])

But Python have the concept of "iterable", meaning some things can be iterated over, without the need of accessing each element individually. For example, our previous list can be iterated with:

a_list = [1, 2, 3, 4]
for value in a_list:
    print(value)

"For every element in a_list, retrieve it and name it value."

A lot of elements are iterable: Strings are iterable, returning every character in them; dictionaries are iterable, returning every key in them; sets are iterable, returning every element in them; tuples are iterable, returning every value in them; generators are iterable, return the next value they can produce.

But what if you need to iterate over more than one iterable at the same time?

Enters zip()

That's where zip() comes in. zip() allows you to merge iterables:

a_list = [1, 2, 3, 4]
a_tuple = ('a', 'b', 'c', 'd')
for mixed_tuple in zip(a_list, a_tuple):
    print(mixed_tuple)

This code prints out:

(1, 'a')
(2, 'b')
(3, 'c')
(4, 'd')

What zip() does is create a tuple with the first element of the first iterable and the first element of the second iterable; then the second element of the first iterable and the second element of the second iterable; and so on. You can put as many iterables as you want in zip() and it will just create larger tuples for each interaction.

Interlude: Destructuring

One of the cool things in Python is "destructuring". Destructuring (de-structuring or more like "breaking apart a structure") allows one to extract elements from a iterable directly.

For example, if you have a tuple with two elements:

a_tuple = (1, 2)

... you'd probably take every element of it in separate variables with something like

a = a_tuple[0]
b = a_tuple[1]

But with destructuring, you can do this in a single pass with

(a, b) = a_tuple

This code and the one above it will do exactly the same thing.

But why destructuring is important if we are talking about iterating over elements? 'Cause for also has the destructuring capabilities:

a_list = [1, 2, 3, 4]
a_tuple = ('b', 'c', 'd', 'f')
a_string = 'aeio'

for (a_number, lowercase_char, uppercase_char) in zip(a_list, a_tuple, a_string):
    print(a_number)
    print(lowercase_char)
    print(uppercase_char)
    print()

{% note() %} Remember that I said that strings are also iterables and each iteration would return a character? That's it. {% end %}

But what happens when one of the iterables is smaller than the other one?

a_short_list = [1, 2]
a_long_list [10, 20, 30, 40, 50, 60, 70, 80, 90]
for (small, big) in zip(a_short_list, a_long_list):
    print(small, big)

That will print

1 10
2 20

zip() stops when the shortest iterable have no more elements. To go as far as the longest iterable, you need itertools.zip_longest().

itertools.zip_longest()

zip_longest(), part of the itertools module, will transverse the iterables till every one of them have no more elements. What happens with the shortest of those is that its value will be replaced with None. Using our previous example:

import itertools

a_short_list = [1, 2]
a_long_list [10, 20, 30, 40, 50, 60, 70, 80, 90]
for (small, big) in itertools.zip_longest(a_short_list, a_long_list):
    print(small, big)

That will print:

1 10
2 20
None 30
None 40
None 50
None 60
None 70
None 80
None 90

Careful with generators

One thing you must be careful when using zip() and zip_longest() are generators. Why? Because some of them have no end.

Let's take one example: cycle(). cycle(), also part of the itertools module, is a generator that, on request, returns the next element of an iterable but, as soon as this iterable is over, it starts over. For example (and I'm tacking zip() around this just for the sake of staying on topic, and you don't need to use zip() with cycle()):

a_list = [10, 20, 30, 40, 50, 60, 70, 80, 90]
for (bullet, value) in zip(cycle(['-', '*', '.']), a_list):
    print(bullet, value)

That will produce:

- 10
* 20
. 30
- 40
* 50
. 60
- 70
* 80
. 90

What happened here is that zip() took the first value of the first iterable, our cycle(['-', '*', '.']), which was the first value of its iterable, '-', and the second value of the second iterable, 10; next iteration, the second value of cycle() was '*' and the second value of a_list was 20; third iteration, cycle() returned '.' and a_list returned 30; now, on the fourth iteration, cycle() was asked for a value and, with its iterable exhausted, it returned to the first value, returning '-' again.

Ok, cool?

So, what's the problem with generators?

Some generators -- like cycle() above -- do not have an end. If you replace zip() with zip_longest() on the code above, you'll see that the code will never stop. It's not every generator the can produce values continuously, though, so you can mess with them with no issue.

{% note() %} It's not zip_longest() that may have an issue. You can put two cycle()s in a zip() and it will keep producing tuples with no end. {% end %}

All nice and dandy, but what if I need to show the index itself?

enumerate() to the rescue!

Ok, so we talked about mixing more than one iterable, but what if we need the position? What if we have a list of ordered results and we need to show the position itself?

Again, you may be temped to use range():

winners = ['first place', 'second place', 'third place', 'fourth place']
for pos in range(len(winners)):
    print(pos + 1, winners[pos].capitalize())

That will print:

1 First place
2 Second place
3 Third place
4 Fourth place

One may also try to be clever and mix our newly found knowledge about zip() and do:

winners = ['first place', 'second place', 'third place', 'fourth place']
for (pos, name) in zip(range(len(winners)), winners):
    print(pos + 1, name.capitalize())

... which ,personally, looks even more cumbersome than the first option. But Python have another generator called enumerate() that takes one single iterable, but produces tuples with the index of it and its value:

winners = ['first place', 'second place', 'third place', 'fourth place']
for (pos, name) in enumerate(winners):
    print(pos + 1, name.capitalize())

Even better, enumerate() have an option to define with will be the value of the first element, so instead of that pos + 1 in the print() statement, we can replace the enumerate to enumerate(winners, start=1) and remove the addition in print().

Conclusion

Iterables is one of the powerhouses of Python, as you may have noticed in the beginning with the number of things that can be iterated over. Understanding those will help you write better and more concise Python code, without losing meaning.