Grouping by key with Python

Today I had to process some data, which was inside an unordered list, using the Python language.

Some computations employed all the list items, others were based only on related ones.

The data was arranged in tuples, each one contained a main value among others.

That value (let's call it "key"), identified a relation with other tuples.

Simplifying it, quite a bit, was something similar to this:

items = [(1, "a"), (3, "q"), (2, "c"), (2, "x"), (1, "z")]

The problem could be solved using some nested "while" iterations.

But actually, I wanted something more brief and readable.

Then, I looked for an alternative, and found itertools.

First I've loaded the required module:

from itertools import groupby

Then, the final solution was much like this:

for key, group in groupby(sorted(items), lambda x: x[0]):
    # Do something with the key
    print(key)
    for tuple in group:
        # Process each tuple with same key
        print(tuple)
    # Other statements
    print("^^^^^^^^^")

There are some remarkable points in this code:

• The list needs to be sorted previously, so keys can be matched up.
• I've used a lambda in order to select which element is the key.
• The key can be accessed individually.
• Related tuples can accessed via the group variable.
• Each item inside each group still contains the key.

This code would print something like this:

1
(1, 'a')
(1, 'z')
^^^^^^^^^
2
(2, 'c')
(2, 'x')
^^^^^^^^^
3
(3, 'q')
^^^^^^^^^

At the end, this method was cleaner than using iterations by hand.