[BioC] finding cliques with packages Graph and RBGL

[lilong at isb-sib.ch]

Hi,

I took a look at the graphs you sent (network{2,3,4}.txt).  I think the
problem has more to do with output format than with the result.  I used
"sapply" instead of "lapply" when converting node index to node names,
which  causes reformatting...

Could you please try to modify maxClique in RBGL/R/interfaces.R to use
"lapply" instead of "sapply" and look at the result?  Let me know if it
works.  If so, I'll check in the fix.

Thanks

Li

> Hi Li,
>
> thanks for your answer.
> I cannot send you the full dataset, because it is a MySQL database of
> various MBs of size, but in the file "data.zip" in attach you find four
> files with the adjacency matrix (NOTE: for a direct graph, so, for
> example,
> 40 19 means there is a link between node 40 and node 19) for package
> "Graph"
> in four temporal steps (note: these steps are consecutive) where I found
> what I described in my first e-mail (but it also happens in other
> intervals).
> The first file ("network1.txt"), maxclique found cliques of length 19, in
> the second and third file ("Network2" and "network3") the maximum clique
> found has length 1. In the fourth file ("network4") length 19.
> I also send you  the plot of some network indicator:
> network density ("densita.png")
> number of active node ("numero_agenti_vivi.png")
> The total step are 500; the adjacency matrices I sent you are relative at
> steps 429-430-431-432
> Thanks in advance
>
> Davide
>
>
> 2006/12/18, lilong at isb-sib.ch <lilong at isb-sib.ch>:
>>
>>
>> Hi,
>>
>> Could you provide the dataset?  Without any further info, it's hard to
>> judge the results, especially the graph changes dynamically.  For
>> instance
>> it's possible that at SOME stage, there is no bigger clicks.
>>
>> If you provide a graph that shows the behaviour (that particular graph),
>> we could help further.
>>
>> Thanks
>>
>> Li
>>
>> > Hello everyone,
>> >
>> > I am using packages Graph and RBGL 1.8 with R 2.3.1 for finding
>> cliques
>> in
>> > a
>> > undirected dynamical graph (it changes over time).
>> > What I do: I load data (adjacency matrix for Graph) from database in a
>> > temporal step.
>> > I transform the matrix in a direct graph using ftM2graphNEL, I make
>> some
>> > analysis (like "degree"), then I transform the
>> > direct graph in a undirect graph using "ugraph" to use maxClique to
>> find
>> > all
>> > the cliques in this graph (which now is undirect).
>> > But I found a problem: in some steps the output of maxClique give me
>> only
>> > clique of length one...but it is a graph with density
>> > near at 1! In the step before it gives me one or two cliques of lenght
>> 27
>> > or
>> > 28 (the node in the graph are 29).
>> > I give you some outputs:
>> >
>> > step "Before":
>> >
>> > $maxCliques
>> > $maxCliques[[1]]
>> >  [1] "1"  "3"  "6"  "7"  "8"  "9"  "11" "12" "14" "16" "17" "18" "19"
>> "20"
>> > "21"
>> > [16] "24" "25" "27" "30" "31" "32" "35" "36" "37" "39" "40" "33" "26"
>> >
>> > $maxCliques[[2]]
>> >  [1] "1"  "3"  "6"  "7"  "8"  "9"  "11" "12" "14" "16" "17" "18" "19"
>> "20"
>> > "21"
>> > [16] "24" "25" "27" "30" "31" "32" "35" "36" "37" "39" "40" "22"
>> >
>> > the degree (in the undirected graph) of each node is (node are number
>> from
>> > 1
>> > to 40, but in time some node (e.g.  node 2) "dies"):
>> >
>> >  1  3  6  7  8  9 11 12 14 16 17 18 19 20 21 22 24 25 26 27 30 31 32
>> 33
>> 35
>> > 36
>> > 28 28 28 28 28 28 28 28 28 28 28 28 28 28 28 26 28 28 27 28 28 28 28
>> 27
>> 28
>> > 28
>> > 37 39 40
>> > 28 28 28
>> >
>> > and in the "next" step the output is the following:
>> >
>> > $maxCliques
>> >       [,1] [,2]
>> >  [1,] "1"  "1"
>> >  [2,] "3"  "3"
>> >  [3,] "6"  "6"
>> >  [4,] "7"  "7"
>> >  [5,] "8"  "8"
>> >  [6,] "9"  "9"
>> >  [7,] "11" "11"
>> >  [8,] "12" "12"
>> >  [9,] "14" "14"
>> > [10,] "16" "16"
>> > [11,] "17" "17"
>> > [12,] "18" "18"
>> > [13,] "19" "19"
>> > [14,] "20" "20"
>> > [15,] "21" "21"
>> > [16,] "24" "24"
>> > [17,] "25" "25"
>> > [18,] "26" "26"
>> > [19,] "27" "27"
>> > [20,] "30" "30"
>> > [21,] "31" "31"
>> > [22,] "32" "32"
>> > [23,] "35" "35"
>> > [24,] "36" "36"
>> > [25,] "37" "37"
>> > [26,] "39" "39"
>> > [27,] "40" "40"
>> > [28,] "33" "22"
>> >
>> > The number of node in each cliques in always 1.
>> >
>> > degree:
>> >  1  3  6  7  8  9 11 12 14 16 17 18 19 20 21 22 24 25 26 27 30 31 32
>> 33
>> 35
>> > 36
>> > 28 28 28 28 28 28 28 28 28 28 28 28 28 28 28 27 28 28 28 28 28 28 28
>> 27
>> 28
>> > 28
>> > 37 39 40
>> > 28 28 28
>> >
>> > Some ideas?
>> > Thank-you very much.
>> >
>> > Sorry for my english.
>> >
>> > Davide
>> >
>> >       [[alternative HTML version deleted]]
>> >
>> > _______________________________________________
>> > Bioconductor mailing list
>> > Bioconductor at stat.math.ethz.ch
>> > https://stat.ethz.ch/mailman/listinfo/bioconductor
>> > Search the archives:
>> > http://news.gmane.org/gmane.science.biology.informatics.conductor
>> >
>>
>>
>