I'm very interested in using cosparallel for my research, and first thank
you very much for inventing this great tool. I don't see an active
discussion or issue tracker for the software so please point me in the
right direction if there is a better venue for this question. I have
attached a sample of output for my graph generated by a command like:
$ ./cos -P 1 input.mcliques
This was a test using one thread. I don't know if I understand the output
and I have not found the answer in the documentation. I hope you can tell
me if this is sensible output, a bug, or if there is additional
post-processing that is required or recommended. I've attached a sample of
output for k=9 communities, as well as the console output. My questions are:
Why are there multiple lines corresponding to the same community id?
The User Guide says about the <community_id>, "This integer uniquely
identifies the community".
Lines with the same <community_id> often contain the same
<maximal_clique> ids. Should these lines be combined and the set of unique
<maximal_clique> ids be treated as the k-clique community?
Are these redundant lines/ids actually correct or do they indicate an
error?
Thank you in advance,
Brooks Ambrose
Brooks Ambrose, M.A., C.Phil.
Instructor, Berkeley School of Information
University of California, Los Angeles
brooksambrose@berkeley.edu
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Each community file (e.g., 9_communities.txt) has multiple lines corresponding to the same community id since each line correspond to a unique maximal clique. By construction, each k-clique community is made up of overplapping maximal cliques but is defined on edges -- and not on nodes. In other words, if you have a k-clique community C spanning a set of nodes N in a graph G, this does not equal the subgraph of G with the nodes in N. For this reason, the software COS outputs the full list of maximal cliques for each community, one per line.
So for the sake of example, if we consider an excerpt of the file 9_communities.txt:
we have that 9-clique community identified by 0 has three overlapping maximal cliques with size 10, among others.
Maximal cliques, expressed al lists of their memers, are indicated on the right of the semi-colon, See for example the overlapping node 2136 in bold above.
To sum up, there are no redundant lines or errors. Everything looks OK.
Hello,
I'm very interested in using cosparallel for my research, and first thank
you very much for inventing this great tool. I don't see an active
discussion or issue tracker for the software so please point me in the
right direction if there is a better venue for this question. I have
attached a sample of output for my graph generated by a command like:
$ ./cos -P 1 input.mcliques
This was a test using one thread. I don't know if I understand the output
and I have not found the answer in the documentation. I hope you can tell
me if this is sensible output, a bug, or if there is additional
post-processing that is required or recommended. I've attached a sample of
output for k=9 communities, as well as the console output. My questions are:
The User Guide says about the <community_id>, "This integer uniquely
identifies the community".
<maximal_clique> ids. Should these lines be combined and the set of unique
<maximal_clique> ids be treated as the k-clique community?
error?
Thank you in advance,
Brooks Ambrose
Brooks Ambrose, M.A., C.Phil.
Instructor, Berkeley School of Information
University of California, Los Angeles
brooksambrose@berkeley.edu
Brooks,
Thanks for reaching out.
Each community file (e.g., 9_communities.txt) has multiple lines corresponding to the same community id since each line correspond to a unique maximal clique. By construction, each k-clique community is made up of overplapping maximal cliques but is defined on edges -- and not on nodes. In other words, if you have a k-clique community C spanning a set of nodes N in a graph G, this does not equal the subgraph of G with the nodes in N. For this reason, the software COS outputs the full list of maximal cliques for each community, one per line.
So for the sake of example, if we consider an excerpt of the file 9_communities.txt:
[...]
0:1042 1587 1590 1594 1598 1607 1613 2134 2136 2299
0:1042 1590 1593 1594 1598 1607 1613 2136 2283 2299
0:1590 1593 1594 1598 1607 1613 1619 2136 2283 2299
[...]
we have that 9-clique community identified by 0 has three overlapping maximal cliques with size 10, among others.
Maximal cliques, expressed al lists of their memers, are indicated on the right of the semi-colon, See for example the overlapping node 2136 in bold above.
To sum up, there are no redundant lines or errors. Everything looks OK.
Simone
ps. I enclose here your files for completeness.