Efficient unique list in Python

Is that even a thing?

Unless you're not totally familiar with Python language, you probably know the answer already. Yes, the holy grail of Python, the list comprehension!

src_list = ['A','B','C','D','E','common_A','common_B']
conditions = ['common_A','common_B']
taget = [el for el in src_list if el in conditions]

Of course, it is.

Let's tweak a bit.

Let's say one of or both lists contain at least 1 million elements. Then, would you still use this method? Especially if you need to manipulate your pandas DataFrame, which has multiple columns to work with, what would you say?

Set is here to help, right?

src_list = ['A','B','C','D','E',..., 'common_A','common_B',...]
conditions = ['common_A','common_B',...]
taget = set(src_list) - set(conditions)

Well, maybe. It's the simplest solution but this high-performing set type is known to be unordered. If the original list is unordered and you need to preserve the ordering, you're in a big trouble.

Dict is here to the rescue!

Another high-performing data type, dict can achieve this additional layer of requirement. It's fast and don't mess with the ordering of the input.

dict.fromkeys(list) is the magic command for you. As the syntax implies, it constructs a dict object from the input (list) and value items are set to None (only key items in the order of presence in the input are added). For the list ['A','B','C','D','E'], this command yields

{'A':None,'B':None,'C':None,'D':None,'E':None}

To apply to the original problem,

src_list = ['A','B','C','D','E',..., 'common_A','common_B',...]
conditions = ['common_A','common_B',...]
taget = [el for el in dict.fromkeys(src_list) if el in set(conditions)]

As an added bonus, we get the unique elements.