Menu
▾
▴
Re: [Jung-support] DegreeDistributionRanker
|
From: Joshua O'M. <jos...@gm...> - 2007-04-17 22:59:58
|
Bea: Anecdotal evidence suggests that migrating to JUNG2 is actually pretty easy (and results in cleaner code), but it's up to you when you want to do that. (Bear in mind that JUNG 1.x will not be improved further at this point, though.) To address your second question: if the graphs are not strongly connected (in the case of directed graphs), the algorithms will not work, as you've seen. Furthermore, each vertex needs to have at least one outgoing edge. This is why the authority and hub sums are < 1 (they should each be 1). Hope this helps-- Joshua On 4/17/07, Beatriz Sevilla <bea...@gm...> wrote: > Thanks a lot!!!!!!! In my next project i will use Jung2 ;P > > I have another question about Hits o Hits Priors, I running with > BarabasiAlbert free-scale nets with 15 nodes and 7 edges per vertex more > or less, no parallel edges and i try undirected and directed graphs. The > case is that most of the time I get : > > HITS With Priors scores can not be generrated because the specified graph is > not connected. > Authority Sum: 0.5900456336154054 > HITS With Priors scores can not be generrated because the specified graph is > not connected. > Hub Sum: 0.5925901779146368 > > But the graphs are connected, at least what I know as connected. > > Thanks again, bea. > > > On 4/17/07, Joshua O'Madadhain <jos...@gm...> wrote: > > Bea: > > > > Ah, now I understand your problem. The short form is that the > > existing graph implementations (in JUNG 1.x) of getVertices() (and the > > other get() methods) do not guarantee an order of iteration. That is, > > if you get an Iterator for getVertices(), it will give you a > > particular ordering, and if you get another Iterator for the _same > > set_ it may give you a completely different ordering. This is > > basically because the Java Set interface provides no guarantees about > > ordering. > > > > In order to get around this, you basically need a different internal > > data structure that hangs onto the vertices. (For example, if you > > replace a HashMap with a LinkedHashMap, the latter has an order of > > iteration that is guaranteed to be the order in which the elements are > > added.) > > > > In the JUNG 2.0 alpha, we've included graph implementations that > > provide ordering guarantees for their various collections. > > > > Joshua > > > > On 4/17/07, Beatriz Sevilla <bea...@gm...> wrote: > > > > > > Sorry if you 've misunderstood my question. I mean, i'm no t judging > > > anything. But I run this algorithm several times with the same graph but > > > different weights and the results changes, and when iI've said are > > > different I mean the order of the nodes with the same number of > connexions, > > > and I thought that maybe there's a reason behind. > > > > > > Nevermind, thank you for answer so fast and overall for this api that is > > > making my project faster ;P > > > Best, bea. > > > > > > > > > On 4/17/07, Joshua O'Madadhain <jos...@gm...> wrote: > > > > Beatriz: > > > > > > > > I'm not sure that I understand what your question is; it sounds like > > > > you're not satisfied with the results of our algorithm, but I'm not > > > > sure what you want. > > > > > > > > > > You've already noticed that our ranking algorithm works on the basis > > > > of degree rather than weighted degree. It sounds like you might want > > > > something that ranks vertices of the same degree according to their > > > > weighted degree. Is that correct? > > > > > > > > Maybe to clarify things you could reorder the output you showed us > > > > (just using cut-and-paste) to show us what order you'd like the > > > > results to be in. > > > > > > > > Joshua > > > > > > > > On 4/17/07, Beatriz Sevilla <bea...@gm...> wrote: > > > > > Hi, > > > > > I'm analysing some social networks, and I have some wierd results on > > > this > > > > > algorithm, so if anyone could help me to understand?? > > > > > My Results: > > > > > > > > > > 1. 8 conexions > > > > > 2. 7 conexions and the sum od weights = 52 > > > > > 3. 7 conexions and the sum od weights = 37 > > > > > 4. 6 conexions and the sum od weights = 23 > > > > > 5. 5 conexions and the sum od weights = 21 > > > > > 6. 4 conexions and the sum od weights = 11 > > > > > 7. 3 conexions and the sum od weights = 16 > > > > > 8. 3 conexions and the sum od weights = 8 > > > > > 9. 3 conexions and the sum od weights = 16 > > > > > 10. 3 conexions and the sum od weights = 15 > > > > > 11. 3 conexions and the sum od weights = 17 > > > > > 12. 3 conexions and the sum od weights = 5 > > > > > 13. 3 conexions and the sum od weights = 8 > > > > > 14. 3 conexions and the sum od weights = 13 > > > > > 15. 3 conexions and the sum od weights = 16 > > > > > > > > > > I understand that is ranking by the number od connexions, but i > don't > > > know > > > > > how to rank when the nodes has the same number of conexions. > > > > > > > > > > Thanks, > > > > > Bea. > > > > > > > > > > -- > > > > > "There are known knowns. There are things we know we know. > > > > > We also know there are known unknowns. That is to say, we > > > > > know there are some things we do not know. But there are > > > > > also unknown unknowns, the ones we don't know we don't know." > > > > > (Donald Rumsfeld) > > > > > > > > > ------------------------------------------------------------------------- > > > > > This SF.net email is sponsored by DB2 Express > > > > > Download DB2 Express C - the FREE version of DB2 express and take > > > > > control of your XML. No limits. Just data. Click to get it now. > > > > > http://sourceforge.net/powerbar/db2/ > > > > > _______________________________________________ > > > > > Jung-support mailing list > > > > > Jun...@li... > > > > > > > > > https://lists.sourceforge.net/lists/listinfo/jung-support > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > jos...@gm......................www.ics.uci.edu/~jmadden > > > > Joshua O'Madadhain: Information Scientist, Musician, > > > Philosopher-At-Tall > > > > It's that moment of dawning comprehension that I live for. -- Bill > > > Watterson > > > > My opinions are too rational and insightful to be those of any > > > organization. > > > > > > > > > > > > > > > > -- > > > > > > "There are known knowns. There are things we know we know. > > > We also know there are known unknowns. That is to say, we > > > know there are some things we do not know. But there are > > > also unknown unknowns, the ones we don't know we don't know." > > > (Donald Rumsfeld) > > > > > > -- > > > jos...@gm......................www.ics.uci.edu/~jmadden > > Joshua O'Madadhain: Information Scientist, Musician, > Philosopher-At-Tall > > It's that moment of dawning comprehension that I live for. -- Bill > Watterson > > My opinions are too rational and insightful to be those of any > organization. > > > > > > -- > > "There are known knowns. There are things we know we know. > We also know there are known unknowns. That is to say, we > know there are some things we do not know. But there are > also unknown unknowns, the ones we don't know we don't know." > (Donald Rumsfeld) -- jos...@gm......................www.ics.uci.edu/~jmadden Joshua O'Madadhain: Information Scientist, Musician, Philosopher-At-Tall It's that moment of dawning comprehension that I live for. -- Bill Watterson My opinions are too rational and insightful to be those of any organization. |