Python: Detect most similar list from list of lists

I want to detect the most similar list from list of lists in the fastest way.

My searching list:

[1,2,3,4]

The list of lists:

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

Most simillar result:

[1,2,3]

I was trying to find that with some common operators in python but it’s too slow in my data. I have about 2 million list of lists that I want to search in them.

>Solution :

The following fonction returns the most similar lists according to the length

def most_similar_acc_length(my_list, range_of_lists, length_range):
    """most similar series according to length
    Parameters
    ----------
    my_list :       The list of interest
    range_of_lists: List of lists where we search the most similar to 'my_list'
    length_range:   Range of series length to be considered as similar to the one of my_list
    
    Returns:
    --------
    List of most similar lists in terms of length
    """
   
    sim_lists=[x for x in range_of_lists if len(x)>=(len(my_list)-length_range) and len(x)<=(len(my_list)+length_range)]
    return sim_lists

If we try it on the lists you shared with length_range length_range=1 we get:

range_of_lists=[[1],[2],[1,2],[1,2,3,4,5,6],[1,2,3],[1,2,3,4,5]]
my_list=[1,2,3,4]

sim_list=most_similar_acc_length(my_list, range_of_lists, 1)

Output

[[1, 2, 3], [1, 2, 3, 4, 5]]

Second step
We set up another function after having similar lists according to length

def most_similar_list(my_list, range_of_lists, length_range):
    # We start with a first selection similar lists in terms of length
    sim_list=most_similar_acc_length(my_list, range_of_lists, length_range)
    
    new_list=[]      # Binary values ==1 when value is same and ==0 when not
    temp_list=[]     # Temprary list to be appended to 'new_list'
    
    for x in sim_list:
        for i in range(min(len(x), len(my_list))):
            if i==min(len(x)-1, len(my_list)-1):
                if x[i]==my_list[i]:
                    temp_list.append(1)
                else:
                    temp_list.append(0)
                new_list.append(temp_list)
                temp_list=[]
            else:
                if x[i]==my_list[i]:
                    temp_list.append(1)
                else:
                    temp_list.append(0)

    max_list=[sum(x) for x in new_list]
    ind_max=max_list.index(max(max_list))
    
    return sim_list[ind_max]

Let’s try this function:

range_of_lists=[[1],[2],[1,2],[1,2,3,4,5,6],[1,2,3],[1,2,3,4,5]]
my_list=[1,2,3,4]

similar_list=most_similar_list(my_list, range_of_lists, 1)

similar_list

Output

[1, 2, 3, 4, 5]