Hello,
I have assembled a network with 10 nodes and 81 directed links to check for collaboration. I was trying to compute the eigenvector centrality values for the different nodes/students collaboration. However, it asks me to input the number of iterations.
I was wondering what number I should input? Also, what does the number of iterations really do? The visualization that I get after is a graph with just 2 data points and hence I don't think it is pulling the data from the data lab that I have created. How can I change this?
Any help would be much appreciated!
Akash
Eigenvector Centrality
-
akash90gupta
- Posts:3
- Joined:04 Aug 2013 00:34
- Location:Boston [phpBB Debug] PHP Warning: in file [ROOT]/vendor/twig/twig/lib/Twig/Extension/Core.php on line 1275: count(): Parameter must be an array or an object that implements Countable
-
pegerp
- Posts:124
- Joined:21 Dec 2011 17:10 [phpBB Debug] PHP Warning: in file [ROOT]/vendor/twig/twig/lib/Twig/Extension/Core.php on line 1275: count(): Parameter must be an array or an object that implements Countable
Re: Eigenvector Centrality
Hi Akash.
The eigenvector centrality algorithm is explained quite well here: http://stackoverflow.com/questions/1445 ... pseudocode (iterations is the value 100 in the example) . The more there are iterations the more accurate the centrality value becomes. In practice value like 100 is good.
For network with only 10 nodes it is possible that Gephi's scatter plot doesn't show that much useful information. It is possible that many nodes share the same centrality value. Most likely the algorithm is still working properly.
After computing the centrality see Data Laboratory > Nodes > Eigenvector Centrality column. These are the values that should interest you.
Using centrality as ranking parameter for node size is often quite good visually. Overview > Ranking > Diamond shape > choose Eigenvector Centrality from the dropdown menu > press Apply.
The eigenvector centrality algorithm is explained quite well here: http://stackoverflow.com/questions/1445 ... pseudocode (iterations is the value 100 in the example) . The more there are iterations the more accurate the centrality value becomes. In practice value like 100 is good.
For network with only 10 nodes it is possible that Gephi's scatter plot doesn't show that much useful information. It is possible that many nodes share the same centrality value. Most likely the algorithm is still working properly.
After computing the centrality see Data Laboratory > Nodes > Eigenvector Centrality column. These are the values that should interest you.
Using centrality as ranking parameter for node size is often quite good visually. Overview > Ranking > Diamond shape > choose Eigenvector Centrality from the dropdown menu > press Apply.
-
shooter61
- Posts:4
- Joined:04 Nov 2013 15:08 [phpBB Debug] PHP Warning: in file [ROOT]/vendor/twig/twig/lib/Twig/Extension/Core.php on line 1275: count(): Parameter must be an array or an object that implements Countable
Re: Eigenvector Centrality
Eigenvector centrality iteration amount is a little bit confusing for me.
I have a network with 255 nodes and 234 edges. When I applied 100 iterations, 10000 iterations and 100000 iterations each time the values change. More importantly, the ranking of the nodes also change.
What is the reason behind this situation? Could somebody explain it.
Moreover, the ties are weighted in this network. Does the software consider these weights while measuring eigenvector centrality?
I have a network with 255 nodes and 234 edges. When I applied 100 iterations, 10000 iterations and 100000 iterations each time the values change. More importantly, the ranking of the nodes also change.
What is the reason behind this situation? Could somebody explain it.
Moreover, the ties are weighted in this network. Does the software consider these weights while measuring eigenvector centrality?