question about edge weights using scluster
Hello I have been experimenting with the different clustering options provided by CLUTO and greatly appreciate the software. I had a few questions when using scluster -clmethod=graph. I have a a problem with well separated clusters but when I try to run the graph techniques they usually take a very long time versus a few seconds for the other options. I suspect this is because I have been setting the -grmodel to one of the sd,ad,sl,al options and in the CLUTO output I see a number of the following warnings
The weight of edge (0 404) is zero. The edge-weights must be greater than zero.
I guess my questions would be as follows:
1 why should the edge weights have to be greater than zeros, doesn't a zero just indicate no connection between nodes?
2 is this really a problem because I am not using the -grmodel=none option
3 I get more than 8 clusters which I in another post should be fixed by increasing -nnbrs
Below I have included more of the CLUTO information. Thanks for any help.
Brendan
********************************************************************************
scluster (CLUTO 2.1.1) Copyright 2001-03, Regents of the University of Minnesota
Graph Information ------------------------------------------------------------
Name: data\in\datatype=i5sim_t=2_simtype=hu_sigma=0.2.sfile, #Vtxs: 800, #Edges: 640000
Options ----------------------------------------------------------------------
CLMethod=GRAPH, CRfun=Cut, #Clusters: 8
EdgePrune=-1.00, VtxPrune=-1.00, GrModel=SY-DIR, NNbrs=400, MinComponent=5
CSType=Best, AggloFrom=24, AggloCRFun=E1, NTrials=1, NIter=10
Solution ---------------------------------------------------------------------
The weight of edge (0 1) is zero. The edge-weights must be greater than zero.
The weight of edge (0 404) is zero. The edge-weights must be greater than zero.
...
...
The weight of edge (798 796) is zero. The edge-weights must be greater than zero.
The weight of edge (798 797) is zero. The edge-weights must be greater than zero.
A total of 142278 errors exist in the input file. Correct them, and run again!
--------------------------------------------------------------------------------------------
10-way clustering: [Cut=1.08e+001] [800 of 800], Entropy: 0.000, Purity: 1.000
--------------------------------------------------------------------------------------------
cid Size ISim ISdev ESim ESdev Entpy Purty | 1 2 3 4 5 6 7 8
--------------------------------------------------------------------------------------------
0 100 +0.001 +0.002 +0.000 +0.000 0.000 1.000 | 0 0 0 100 0 0 0 0
1 100 +0.001 +0.001 +0.000 +0.000 0.000 1.000 | 100 0 0 0 0 0 0 0
2 100 +0.000 +0.001 +0.000 +0.000 0.000 1.000 | 0 100 0 0 0 0 0 0
3 41 +0.000 +0.001 +0.000 +0.000 0.000 1.000 | 0 0 0 0 41 0 0 0
4 100 +0.000 +0.000 +0.000 +0.000 0.000 1.000 | 0 0 100 0 0 0 0 0
5 100 +0.000 +0.000 +0.000 +0.000 0.000 1.000 | 0 0 0 0 0 0 0 100
6 1 -1.#IO -1.#IO +0.000 -1.#IO 0.000 1.000 | 0 0 0 0 1 0 0 0
7 100 +0.000 +0.000 +0.000 +0.000 0.000 1.000 | 0 0 0 0 0 0 100 0
8 100 +0.000 +0.001 +0.000 +0.000 0.000 1.000 | 0 0 0 0 0 100 0 0
9 58 +0.000 +0.000 +0.000 +0.000 0.000 1.000 | 0 0 0 0 58 0 0 0
--------------------------------------------------------------------------------------------
Timing Information -----------------------------------------------------------
I/O: 0.328 sec
Clustering: 622.984 sec
Reporting: 0.016 sec
********************************************************************************
- Login to post comments
RE: Can you open an issue using
Can you open an issue using the flyspray bug tracking tool and upload the data file there so that I can take a look at it.
RE: Thank you for the quick
Thank you for the quick reply. I submitted the issue to flyspray along with the GraphFile I am using.