Itertools.chain.from_iterable

Can anyone explain to me, what exactly this code snippet is doing? chained_country_list = set(itertools.chain.from_iterable(country_and_countrycodes)) & set(all_countries) I k

Solution 1:

Let's break down each significant element of the code:

itertools.chain.from_iterable:

Basically, this is used to flatten a nested list, like this:

l = [[0], [1, 2], [2], [3, 6], [4], [5, 10]]
list(itertools.chain.from_iterable(l))    

Output:

[0, 1, 2, 2, 3, 6, 4, 5, 10]

& operator between two sets:

Consider the follow example of sets a and b.

a = {1, 2, 3}
b = {2, 3, 4}
a & b

Output:

{2, 3}

So basically it gets the common elements between two sets. Here they're 2 and 3.

The code as a whole:

Let's say:

country_and_countrycodes = [('United States', 'US'), ('China', 'CH')]
all_countries = ['United States', 'Mongolia', 'Togo']

Now, the first part is:

set(itertools.chain.from_iterable(country_and_countrycodes))

which gives us:

{'CH', 'China', 'US', 'United States'}

So, it just gets us a flat set from the tuples.

Then, the second part is:

set(itertools.chain.from_iterable(country_and_countrycodes)) & set(all_countries)

which gives us:

{'United States'}

Basically, what we did was:

{'CH', 'China', 'US', 'United States'} & {'United States', 'Mongolia', 'Togo'}

Since the only common element here is 'United States', that's what we got as the output.

Solution 2:

It's hard to tell what does your code do without knowing the type and value of the objects passed to the functions. However, the function chain.from_iterable tries to create a flattened iterable off of country_and_countrycodes which presumably should be a nested iterables like a nested list. At next step the set function creates a set from the flattened result in order to be and-able with set(all_countries).

Now as a more Pythonic alternative to the following part:

set(itertools.chain.from_iterable(country_and_countrycodes))

You could just pass the iterables to set().union() function in order to create a union set of unique items at once.

Example:

In [2]: set().union(*[[1, 3], [5, 6], [3, 5]])
Out[2]: {1, 3, 5, 6}

So, you can change that code to the following:

set().union(*country_and_countrycodes) & set(all_countries)
# Or
# set().union(*country_and_countrycodes).intersection(all_countries)