merging list of lists

Question

Possible Duplicate:
I want to group tuples based on similar attributes

I am very new to Python and coding in general, have a list of cluster pairs like so

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

I have written this code to join the cluster pairs into list of larger clusters.

for i in range (len(Clist)):
    for j in range(i+1,len(Clist)):
        for disk in Clist[i]:
            if disk in Clist[j]:
                joined=joincluster(Clist[i], Clist[j])
                Clist[i]=list(set(joined))
                Clist[j]=[]
                break

I go on to take the empty ones out. However it doesn't join all the pairs up when i have more then 30 or so cluster pairs, some clusters that should have been joined have not been. I can't see where I'm going wrong? Can anybody me identify the problem?

[[0, 2], [0, 9], [1, 10], [2, 0], [2, 9], [3, 6], [3, 10], [3, 11], [4, 5], [4, 7], [4, 12], [4, 14], [5, 4], [5, 8], [5, 12], [5, 14], [6, 3], [6, 10], [6, 11], [7, 4], [7, 12], [8, 5], [8, 14], [9, 0], [9, 2], [9, 13], [10, 1], [10, 3], [10, 6], [10, 11], [11, 3], [11, 6], [11, 10], [12, 4], [12, 5], [12, 7], [13, 9], [14, 4], [14, 5], [14, 8]]

Sorry, problem only happens when i have a fairly large number of pairs. And this is what I get

[[0, 9, 2, 13], [11, 1, 10, 3, 6], [3, 6], [4, 5, 7, 8, 12, 14]]

so the 3,6 there should have been merged but it hasn't.

How exactly do you want to merge them? We need to see some logic behind your input and output, which clusters do you want to merge with which? — , Jan 19 '13 at 15:13
Well i just want to merge the lists, that contain the same numbers, so for example [1,2] gets merged with [2,8] to make [1,2,8]. Im not sure what you mean by where, but the merged lists should be in the same original list that the clusters were in, if its easier, in a completely new list. — , Jan 19 '13 at 15:16
You need to provide more details on what you actually want to achieve. What are the exact rules? — dzida, Jan 19 '13 at 20:42

score 1 · Answer 1 · 2013-01-19T16:10:31.760

Let's take a simpler example, with only 4 clusters like so:

clusters = [ [1, 10], [1, 20], [3, 6], [3, 10] ]

These 4 are enough to show why [3, 6] gets separated from all other clusters. You loop through these clusters starting from first to last. Here's what happens:

Is there already a cluster that has 1 or 10 in it? - No, create a new one including both. We now have the following: [[1, 10]].
Is there already a cluster that has 1 or 20 in it? - Yes, append to the match. We now have the following: [[1, 10, 20]].
Is there already a cluster that has 3 or 6 in it? - No, create a new one including both. We now have the following: [[1, 10, 20], [3, 6]].
Is there already a cluster that has 3 or 10 in it? - Yes, append to the match. We now have the following: [[1, 3, 10, 20], [3, 6]].

Problem is: after appending [3, 10] to [1, 10], it doesn't check if there are existing clusters that have 3 in them already, which don't have 10. You need to add that to your code and it will work fine.

merging list of lists

1 Answers1