DEV Community

Cover image for 5 Python Tricks in one line. Beginner vs Professional

5 Python Tricks in one line. Beginner vs Professional

Vadim Kolobanov on November 07, 2021

Photo by AltumCode on Unsplash Python is famous for its simplicity and ease of use. This is achieved, among other things, due to the compactness o...
Collapse
 
kbauer profile image
kbauer

There are errors in the part about removing duplicates.

  1. fruits = { ... } already is a set, and so the duplicates are already discarded before reaching the loop.

  2. The output of

    new = set(fruits)
    

    is not

    ['banana', 'apple', 'pear', 'orange']
    

    but (order may differ)

    {'banana', 'apple', 'pear', 'orange'}
    

    The output you show would be obtained by

    new = list(set(fruits))
    

Overall, to be correct, the example would have to be changed to

# List instead of set
fruits = ['apple', 'orange', 'apple', 'pear', 'orange', 'banana']

# Actually make `new` a list
new = list(set(fruits))
Enter fullscreen mode Exit fullscreen mode

Additionally, for a beginner post it would be very useful to point out that, in addition to be more concise,

new = list(set(fruits))
Enter fullscreen mode Exit fullscreen mode

has O(N) time complexity[1], while

for fruit in fruits:
    if fruit not in new:
        new.append(fruit)
Enter fullscreen mode Exit fullscreen mode

has O(N²) time complexity.

[1] In CPython anyway, which should be using a hash-table based set implementation. Note sure though, whether the loop or the set based solution is faster for small numbers of elements.

Collapse
 
ruudvanderham profile image
ruudvanderham

Under Python 3.6 and higher this one liner will also maintain the order
new = list(dict.fromkeys(fruits))
and it is O(n)!

Collapse
 
vadimkolobanov profile image
Vadim Kolobanov

Thank you all so much) You really help all readers to become a little more literate

Collapse
 
kbauer profile image
kbauer

Interesting to know, but hard to read :/ If I'd see this in the code I wouldn't be aware, that preserving the order is an intent here.

Thread Thread
 
ruudvanderham profile image
ruudvanderham

Just adding a comment would help.

Thread Thread
 
vadimkolobanov profile image
Vadim Kolobanov

Guys, I pointed out your help in the article and asked you to pay attention to the comments) So thank you very much! Many opinions, and various practices make this article even better)

Collapse
 
vadimkolobanov profile image
Vadim Kolobanov

Thank you so much for the addition. Initially, my goal was just to show the difference in solutions to some problems. I will definitely add your suggestion about algorithmic complexity! And you also gave me the idea to write an article about calculating algorithmic complexity

Collapse
 
qazzquimby profile image
Toren Darby

This comment is very critical, but is not meant to be hostile.

The objective should be clarity, not brevity. Writing shorter code is good only to the extent that it makes the code more readable. "Tricks" in general tend to be obscure, make the writer feel clever, and confuse the reader. Even something like slicing's step parameter should often be explained when it's used.

I recommend people learning python read Google's python style guide to understand the reasoning, not necessarily to follow. They would not encourage a nested comprehension for product. google.github.io/styleguide/pyguid...

"a real Python professional will only laugh at you." The passages like this bother me.
Comprehensions are not dogmatically correct, and someone laughing at a beginner for not using advanced features should not be the face of "real python professionals".

With that out of the way, I think these are useful things to know.
In general I'd advise someone to learn by googling what they're trying to do (eg python remove duplicates from list), look at top answers, and consider what is clearest.

Collapse
 
doroshenkoviktor profile image
Viktor Doroshenko

So agree. I read this article with only thought - those explicit code, which author claims as junior's often is much more readable than suggested oneliners. And claims like "a real competent Python developer code" e.t.c are controversial for me. My experience reading code of a real highly experienced and clever professionals tells me that they usually tend to the simplest and most explicit solution possible. Their code is easy to read and understand. But tricky things usually appear because of ego and lack of self confidence as a professional. Real professional developer visible by his patterns, architecture and code style, not by some tricks.

Collapse
 
xtofl profile image
xtofl

If you replace the nested for loops (plain or in a generator expression) with a call to itertools.product, the reader will immediately see the intent.

result = product(first, second)
Enter fullscreen mode Exit fullscreen mode

Most of the times, this carthesian product is only iterated over and above example will do. If it needs to be stored, you'll need a tuple:

result  = [tuple(*p) for p in
    product(first, second, third)
]
Enter fullscreen mode Exit fullscreen mode

Notice how each added level requires only an added argument to product.

Collapse
 
jackm767 profile image
jackm767

In a post for beginners it’s probably worth mentioning that sets are unordered. If you use a set to remove duplicates you can no longer index the data. Converting back to a list may not keep the original order either (? unsure about this).

Collapse
 
vadimkolobanov profile image
Vadim Kolobanov

You are absolutely right. It may be worth adding the following

The set does not contain duplicate elements;

The elements of the set are immutable (they cannot be changed), but the set itself is changeable, and it can be changed;

Since the elements are not indexed, the sets do not support any slicing and indexing operations.

Collapse
 
vadimkolobanov profile image
Vadim Kolobanov

I am infinitely glad that I can help anyone) Now my friends and I want to create a training course for guys like you. Give you information from experience and practice with homework and supervision. So far, our plan is to do it for free, so that we can learn by ourselves. If it will be interesting, then I left my contacts under the article. Connect with us we want to help you

Collapse
 
Sloan, the sloth mascot
Comment deleted
 
vadimkolobanov profile image
Vadim Kolobanov

Здравствуйте) для моей идеи вы просто золото) Правда я пока не совсем сформировал конечную задумку но я с удовольствием с вами посотрудничаю. Так как анлийский язык у меня только для чтения документации подходит)

Thread Thread
 
Sloan, the sloth mascot
Comment deleted
 
vadimkolobanov profile image
Vadim Kolobanov

Как вам удобно) быть может вы мне тоже поможете сформировать конечную идею)

 
vadimkolobanov profile image
Vadim Kolobanov

напишите мне в дискорд если не сложно vadimkolobanov#5875

Collapse
 
foxy4096 profile image
Foxy4096

Love this post, it will surely help me.

Collapse
 
vadimkolobanov profile image
Vadim Kolobanov

I am very happy about it! If my work helped at least 1 person, then it was not in vain and my goal was achieved)

Collapse
 
sentayehu profile image
Sent

Removing duplicate in case insensitive list.
eg fruits=['banana', 'Banana', 'Mango', 'mango', 'orange', 'OranGe']
set(map(str.lower, fruits))
will give you: {'orange', 'banana', 'mango'}

Collapse
 
vadimkolobanov profile image
Vadim Kolobanov

It's very simple) You are trying to change the values of the set. You forget that the values in it are immutable, so the lower case will not work. In this situation, it is necessary to bring to one register earlier

Collapse
 
sentayehu profile image
Sent

Right, but set is applied right after we lower the case for all items.
demo: online-python.com/isJQnjgrZf

Collapse
 
aatmaj profile image
Aatmaj

Nice. But I personally prefer the beginner versions.

Collapse
 
hikionori profile image
Danila Verbinsky

It seems to me that all novice developers know about generators, since they are used in all tutorials, and you will not knowingly know how to use them.

Collapse
 
vadimkolobanov profile image
Vadim Kolobanov

if you look at the projects on github, you can see that there are almost never such solutions or generators in the principle. I mean newbie accounts with 1 repository. They know but they don't use it