How do you sort a list by the second (or nth) element on in Python? How can you leave the first or nth elements in a list as they are but then have the remaining elements in the list sorted?
When applying the
.sort()
list method on a list Python will sort
all
elements in that list, but what if you only wanted to sort after a certain index number?
For example, suppose I had a list with the following contents:
>>> my_list = ['z', 'b', 'c', 'a']
And from this list, I wanted to keep the first element,
'z'
, in the first position, but for the other elements in the list, I wanted them sorted. In other words, I wanted the list to produce the result of
['z', 'a', 'b', 'c']
.
Here’s how this can be achieved using a one-liner :
>>> my_list = ['z', 'a', 'b', 'c']
>>> my_list[:1] + sorted(my_list[1:])
['z', 'a', 'b', 'c']
It’s
very important
to use the built-in
sorted()
function rather than the
.sort()
list function as the
.sort()
list function will
mutate
the original list.
So where is this technique actually useful?
I have found working with dictionaries and outputting their result to CSV that this technique comes in very handy .
Here’s an example working with a list of dictionaries and outputting the result to CSV.
List Of Dictionaries To CSV With Ordered Headers
I had a recent need while programming in Python where I needed to sort
after
the
second element onwards
. I was outputting a CSV file from a
List[Dict]
and I wanted the output of the resulting
csv
file to mimic the order of the first few
Dict
values, but then
after
the
second element
in the list I wanted to rearrange the order of the list by sorting it.
Here is a simplified representation of the
Dict
and the resulting
List
:
data = [{
'Surname': "Smith",
'Address': "1 Way Road, TEST",
'2020-Census': 4,
'2016-Census': 3,
'2012-Census': 2
},
{...}]
# get list of keys for csv.DictWriter
keys_list = list(data[0].keys())
# display list
print(keys_list)
# ['Surname', 'Address', '2020-Census', '2016-Census', '2012-Census']
The output desired of the
csv
file was to keep the
Surname
and
Address
columns at the front of the resulting
csv
file, but then for the remaining
20XX-Census
fields to be placed in a specific order.
To order the keys list the way I need, I would have to slice the list to remove the fields I want static, then with other fields I would sort them and then concatenate them together.
Step 1 – Get Sorting List
In my example here, the first two elements in the list I need to remain static, but the rest I need to sort. The easiest way to perform this operation is to slice the list using the slice operator.
If you recall a
List
is made up of elements that have an index, and could be shown using a diagram like so:
| Index Number | 0 | 1 | 2 | 3 | 4 |
|---|---|---|---|---|---|
| List Elements | Surname | Address | 2020-Census | 2016-Census | 2012-Census |
Therefore, to obtain a section of the list for sorting using the slice operator, I need everything after the second element .
To use the slice operator, we need to recall that it’s syntax looks as follows:
List|String[{start}:{stop{:step}}]
The syntax for using the slice operator on a
List
or
String
is that you can enter a starting
index number
in the
start
area, (it is optional, and this is why it is wrapped with
{}
) and this starting index number is inclusive. Therefore, if we were to slice our
List
above
after
the second element, we would enter the slice operator as follows:
keys_list[2:]
# ['2020-Census', '2016-Census', '2012-Census']
As we wanted this list to continue to the end, we didn’t apply any value in the
stop
section of the slice operator. We could have placed the value of
6
in the
stop
value for the slice operator, and this would achieved the same result:
keys_list[2:6]
# ['2020-Census', '2016-Census', '2012-Census']
As I prefer the former example, where no
stop
value was placed, and I want to capture this as a separate list, I then used the following operation:
order_list = keys_list[2:]
print(order_list)
# ['2020-Census', '2016-Census', '2012-Census']Now that I have the list of entries that I want to sort, I can apply a sort on this list.
Step 2 – Sort List
With our sorting list available, we can apply one of two ways to sort: we can mutate the list, or not mutate the list. By mutating the list, the order of the list will be changed permanently. Let’s do that and show the result:
order_list.sort()
print(order_list)
# ['2012-Census', '2016-Census', '2020-Census']
By applying the
.sort()
list method on our
order_list
variable, the order of the list has now permanently changed.
Step 3 – Concatenate List
Finally, to put this list back together to form the
fieldnames
parameter needed for the
csv.DictWriter
function, we simply apply the
+
operator to both lists:
new_keys_list = keys_list[:2] + order_list
print(new_keys_list)
# ['Surname', 'Address', '2012-Census', '2016-Census', '2020-Census']One Liner
Or, if you wanted to perform a sort
after
the second element
you could do it as a
one-liner
(using the
sorted
function which
does
not
mutate
the original list):
sorted_list = keys_list[:2] + sorted(keys_list[2:])
print(sorted_list)
# ['Surname', 'Address', '2012-Census', '2016-Census', '2020-Census']Or, to turn it into something useable if you find this becomes a common function in your code, but instead of it being the second element it was something different like the nth element , we could write a function as follows:
from typing import List
def sorted_list_from_nth_element(l: List, n: int, reverse: bool = False) -> List:
"""
:param l: list to obtain elements from
:param n: nth element to keep static, same as index number of first element to sort from
:param reverse: default is ascending order, if descending order needed set to True
"""
return l[:n] + sorted(l[n:], reverse=reverse)
# noticed I've jumbled the order here to demonstrate, ascending & descending
key_list = ['Surname', 'Address', '2016-Census', '2020-Census', '2012-Census']
# ascending list
asc_list = sorted_list_from_nth_element(key_list, 2)
print(asc_list)
# ['Surname', 'Address', '2012-Census', '2016-Census', '2020-Census']
# descending list
desc_list = sorted_list_from_nth_element(key_list, 2, True)
print(desc_list)
# ['Surname', 'Address', '2020-Census', '2016-Census', '2012-Census']
# ...and to show the original list hasn't mutated
print(key_list)
# ['Surname', 'Address', '2016-Census', '2020-Census', '2012-Census']Summary
In this article I discovered how to sort a list in python after the second element onwards which was a requirement I had in a current Python project. I also showed the differences between using the
.sort()
method on a list which mutates the original list, and a non-mutating function
sorted()
that returns a new list and doesn’t mutate the original list.
Once we had sorted the portion of the original list it was then a simple matter of concatenating the non-sorted portion with the sorted portion using the
+
operator.
