How to get the indices of reversed duplicates in a list

February 17, 2022

I have the following list:

list1 =[('a','b'),('c','d'),('e','f'),('g','h'),('b','a'),('e','d'),('e','g'),('h','g')]

I wish to append the indices of reversed duplicates from this list. For example:

('a','b') == ('b','a')
('g','h') == ('h','g')

I tried

exec =[]
for i,x in enumerate(list1):
    z=x[::-1]
    if z in list1:
        exec.append(i)
        list1.remove(z)

I got:

exec
[0, 3]

Which is correct. However, this is very inefficient when running on a 10 million elements of list. I know that I can directly remove the reversed duplicates by:

data = list({tuple(sorted(item)) for item in list1})

But I only want to identify the indices of reversed duplicates here. Is there a better way to do it? Thanks in advance.

>Solution :

You can use tuples as a key of a dictionary, so create a dictionary where the sorted tuples are keys and the values are the first index where that tuple occurs in your list. Then, when you encounter a tuple that’s already in the dictionary, you can add the index to your output list:

indices_dict = {}
dup_indices = []

for index, item in enumerate(list1):
    s_item = tuple(sorted(item))
    if s_item in indices_dict:
        dup_indices.append(indices_dict[s_item])
    else:
        indices_dict[s_item] = index

Which gives