jgrapht-users — JGraphT user list.

[jgrapht-users] Jgrapht - tracking DepthFirstSearch iterator backtracks

[Traven <mar...@gm...>]

Hi, I am writing in regards to this question, that I posted on stackoverflow
and couldn't get any answer:

http://stackoverflow.com/questions/30938461/jgrapht-tracking-depthfirstsearch-iterator-backtracks
<http://stackoverflow.com/questions/30938461/jgrapht-tracking-depthfirstsearch-iterator-backtracks>  

Copy of stack's question content is below:

I am trying to use jgrapht DepthFirstIterator to traverse through graph and
find ANY path leading from start node to the target node.

As far as I understand, iterator traverses from vertex to vertex and when it
cannot go any deeper, it backtracks and tries other paths (depth-first
search algorithm).

I know that you cant iterate through the nodes in this way:

iterator = new DepthFirstIterator<Integer, DefaultEdge>(graph);
    while (iterator.hasNext()) {
        ...
    }

My question is:
1. How in jgrapht save the final path found by this algorithm? I can break
the loop after I've found the target node, although I dont have any good
ideas of how to save the edges leading to the solution. I cannot save every
traversed edge, because it will also add all the backtracks to my result.
2. Also, is the Depth First Search Iterator protected against running into
the infinite loop when encountering cycles?

Any help/suggestions highly appreciated :).



--
View this message in context: http://jgrapht-users.107614.n3.nabble.com/Jgrapht-tracking-DepthFirstSearch-iterator-backtracks-tp4024995.html
Sent from the jgrapht-users mailing list archive at Nabble.com.

Re: [jgrapht-users] add edge to directed graph in both directions

[krutor <mat...@gm...>]

Hi everyone,I ended up using undirected graph (WeightedPseudograph) with my
Edge Implementation that has "directed" flag. I hid this WeightedPseudograph
behind my own interface that knows about mixed nature of the graph and
modifies graph behaviour for its clients (it also deals with multithreading
and other staff...). For example I had to modify computation of In Edges for
vertex to account for edges directed flag like this:
public Set getInEdges(Node node) {    Set inEdges = new HashSet<>();    try
{        readLock();                Set edges = graphImpl.edgesOf(node); //
graphImpl = WeightedPseudograph<Node, Edge>        for (Edge edge : edges) {             
if (!edge.isDirected() || edge.getTarget().equals(node)) {               
inEdges.add(edge);            }        }        return
Collections.unmodifiableSet(inEdges);    } finally {        readUnlock();   
}}
However this way you are not able to use JGraphT algorithms for directed
graphs - for example DFS algorithm now treats the graph as undirected, so it
finds path between target and source node even on directed edge (which is ok
in my case)...I chose this implementation only because I need to
differentiate between case when there is undirected edge between two nodes
and when there is actually two opposite directed edges between them...
Otherwise I would use directed graph and represent undirected edge as two
opposite directed edges, which is IMHO cleaner solution.



--
View this message in context: http://jgrapht-users.107614.n3.nabble.com/add-edge-to-directed-graph-in-both-directions-tp4024989p4024994.html
Sent from the jgrapht-users mailing list archive at Nabble.com.

Re: [jgrapht-users] add edge to directed graph in both directions

[Dimitris M. <dma...@cs...>]

Hi Joris,

I remember that in the last entry on the referred thread 
(http://jgrapht-users.107614.n3.nabble.com/Mixed-graphs-td4024952.html), 
krutor said something about creating his/her own implementation to solve 
this exact problem. It would be awesome if krutor could share the code 
with everyone by contributing to the project's github repository!

Best,
Dimitris


On 23/06/2015 08:27 μμ, Joris Kinable wrote:
> Hm, I remember seeing that thread. The two solutions proposed were:
>
> 1. Create 2 graphs, a directed and an undirected graph, and create a 
> GraphUnion from these two graphs. This approach has 2 major disadvantages:
> -The GraphUnion is read-only, so removing a vertex or edge (which are 
> both well defined operations in 'mixed' graphs) are not supported
> -Several algorithms in JgraphT won't work well. For example, one may 
> be interested in finding all Strongly Connected Components in the 
> graph (which again is well-defined for mixed graphs). 
> The StrongConnectivityInspector class for example in JgraphT expects a 
> DirectedGraph as input, and hence, will not accept the GraphUnion.
>
> 2. To represent an undirected edge e=(i,j) in a directed graph, create 
> 2 edge objects, say e1 and e2, which contain the same information, and 
> add them to the directed graph:
> directedGraph.addEdge(i,j,e1);
> directedGraph.addEdge(j,i,e2);
> This approach has again 2 major disadvantages:
> -e1 and e2 duplicate the same information, thereby causing a lot of 
> memory overhead
> -It may prove very difficult to keep the information in the objects e1 
> and e2 synchronized.
>
> So far, neither of these approaches seem desirable. Any remarks on this?
>
> Joris
>
>
> On Tue, Jun 23, 2015 at 12:52 PM, Szabolcs Besenyei 
> <bes...@gm... <mailto:bes...@gm...>> wrote:
>
>     http://jgrapht-users.107614.n3.nabble.com/Mixed-graphs-td4024952.html
>
>     Üdvözlettel,
>
>     Besenyei Szabolcs
>
>     2015-06-23 18:30 GMT+02:00 Joris Kinable <de...@gm...
>     <mailto:de...@gm...>>:
>
>         Say I have a list of directed and undirected edges which I
>         would like to represent in the same graph. Every undirected
>         edge (i,j) can be represented in a directed graph by adding
>         two arcs: (i,j) and (j,i). The following code does however
>         *not* achieve this:
>
>         public  static void main(String[] args){
>             DirectedGraph<Integer, String> directedGraph=new
>         SimpleDirectedGraph<Integer, String>(String.class);
>             String s="UndirectedEdge";
>             directedGraph.addVertex(1);
>             directedGraph.addVertex(2);
>         System.out.println(directedGraph.addEdge(1,2,s));
>         System.out.println(directedGraph.addEdge(2,1,s));
>         }
>
>         The second addEdge call will return false because edge s is
>         already in the graph. What would be a clean way to resolve
>         this issue?
>
>         Thanks,
>
>         Joris
>
>         ------------------------------------------------------------------------------
>         Monitor 25 network devices or servers for free with OpManager!
>         OpManager is web-based network management software that monitors
>         network devices and physical & virtual servers, alerts via
>         email & sms
>         for fault. Monitor 25 devices for free with no restriction.
>         Download now
>         http://ad.doubleclick.net/ddm/clk/292181274;119417398;o
>         _______________________________________________
>         jgrapht-users mailing list
>         jgr...@li...
>         <mailto:jgr...@li...>
>         https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>
>
>
>
> ------------------------------------------------------------------------------
> Monitor 25 network devices or servers for free with OpManager!
> OpManager is web-based network management software that monitors
> network devices and physical & virtual servers, alerts via email & sms
> for fault. Monitor 25 devices for free with no restriction. Download now
> http://ad.doubleclick.net/ddm/clk/292181274;119417398;o
>
>
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users

Re: [jgrapht-users] add edge to directed graph in both directions

[Daniels, T. (US) <tro...@ba...>]

For option 2, one solution to both issues is to only associate a single piece of data with each edge, which is a reference to a shared object containing all of the information.  Now updating the information of one automatically updates the other.  The memory overhead is two references plus the additional arc.

The case where it will not work is if there are algorithms that build internal data structures that store information about the arc.  Those algorithms will not know that the two arcs are the same and need their data shared.

Troy

From: Joris Kinable [mailto:de...@gm...]
Sent: Tuesday, June 23, 2015 1:27 PM
To: Szabolcs Besenyei
Cc: jgr...@li...
Subject: Re: [jgrapht-users] add edge to directed graph in both directions

*** WARNING ***
This message originates from outside our organization, either from an external partner or the internet. Consider carefully whether you should click on any links, open any attachments or reply. For additional information about Spearphishing, click here<http://intranet.ent.baesystems.com/howwework/security/spotlights/Pages/SpearphishingTips.aspx>. If you feel the email is suspicious, please follow this process.<http://intranet.ent.baesystems.com/howwework/security/spotlights/Documents/Dealing%20With%20Suspicious%20Emails.pdf>

Hm, I remember seeing that thread. The two solutions proposed were:

1. Create 2 graphs, a directed and an undirected graph, and create a GraphUnion from these two graphs. This approach has 2 major disadvantages:
-The GraphUnion is read-only, so removing a vertex or edge (which are both well defined operations in 'mixed' graphs) are not supported
-Several algorithms in JgraphT won't work well. For example, one may be interested in finding all Strongly Connected Components in the graph (which again is well-defined for mixed graphs). The StrongConnectivityInspector class for example in JgraphT expects a DirectedGraph as input, and hence, will not accept the GraphUnion.

2. To represent an undirected edge e=(i,j) in a directed graph, create 2 edge objects, say e1 and e2, which contain the same information, and add them to the directed graph:
directedGraph.addEdge(i,j,e1);
directedGraph.addEdge(j,i,e2);
This approach has again 2 major disadvantages:
-e1 and e2 duplicate the same information, thereby causing a lot of memory overhead
-It may prove very difficult to keep the information in the objects e1 and e2 synchronized.

So far, neither of these approaches seem desirable. Any remarks on this?

Joris


On Tue, Jun 23, 2015 at 12:52 PM, Szabolcs Besenyei <bes...@gm...<mailto:bes...@gm...>> wrote:
http://jgrapht-users.107614.n3.nabble.com/Mixed-graphs-td4024952.html

Üdvözlettel,

Besenyei Szabolcs

2015-06-23 18:30 GMT+02:00 Joris Kinable <de...@gm...<mailto:de...@gm...>>:
Say I have a list of directed and undirected edges which I would like to represent in the same graph. Every undirected edge (i,j) can be represented in a directed graph by adding two arcs: (i,j) and (j,i). The following code does however *not* achieve this:

public  static void main(String[] args){
    DirectedGraph<Integer, String> directedGraph=new SimpleDirectedGraph<Integer, String>(String.class);
    String s="UndirectedEdge";
    directedGraph.addVertex(1);
    directedGraph.addVertex(2);
    System.out.println(directedGraph.addEdge(1,2,s));
    System.out.println(directedGraph.addEdge(2,1,s));
}

The second addEdge call will return false because edge s is already in the graph. What would be a clean way to resolve this issue?

Thanks,

Joris

------------------------------------------------------------------------------
Monitor 25 network devices or servers for free with OpManager!
OpManager is web-based network management software that monitors
network devices and physical & virtual servers, alerts via email & sms
for fault. Monitor 25 devices for free with no restriction. Download now
http://ad.doubleclick.net/ddm/clk/292181274;119417398;o
_______________________________________________
jgrapht-users mailing list
jgr...@li...<mailto:jgr...@li...>
https://lists.sourceforge.net/lists/listinfo/jgrapht-users

Re: [jgrapht-users] add edge to directed graph in both directions

[Joris K. <de...@gm...>]

Hm, I remember seeing that thread. The two solutions proposed were:

1. Create 2 graphs, a directed and an undirected graph, and create a
GraphUnion from these two graphs. This approach has 2 major disadvantages:
-The GraphUnion is read-only, so removing a vertex or edge (which are both
well defined operations in 'mixed' graphs) are not supported
-Several algorithms in JgraphT won't work well. For example, one may be
interested in finding all Strongly Connected Components in the graph (which
again is well-defined for mixed graphs). The StrongConnectivityInspector
class for example in JgraphT expects a DirectedGraph as input, and hence,
will not accept the GraphUnion.

2. To represent an undirected edge e=(i,j) in a directed graph, create 2
edge objects, say e1 and e2, which contain the same information, and add
them to the directed graph:
directedGraph.addEdge(i,j,e1);
directedGraph.addEdge(j,i,e2);
This approach has again 2 major disadvantages:
-e1 and e2 duplicate the same information, thereby causing a lot of memory
overhead
-It may prove very difficult to keep the information in the objects e1 and
e2 synchronized.

So far, neither of these approaches seem desirable. Any remarks on this?

Joris


On Tue, Jun 23, 2015 at 12:52 PM, Szabolcs Besenyei <bes...@gm...>
wrote:

> http://jgrapht-users.107614.n3.nabble.com/Mixed-graphs-td4024952.html
>
> Üdvözlettel,
>
> Besenyei Szabolcs
>
> 2015-06-23 18:30 GMT+02:00 Joris Kinable <de...@gm...>:
>
>> Say I have a list of directed and undirected edges which I would like to
>> represent in the same graph. Every undirected edge (i,j) can be represented
>> in a directed graph by adding two arcs: (i,j) and (j,i). The following code
>> does however *not* achieve this:
>>
>> public  static void main(String[] args){
>>     DirectedGraph<Integer, String> directedGraph=new
>> SimpleDirectedGraph<Integer, String>(String.class);
>>     String s="UndirectedEdge";
>>     directedGraph.addVertex(1);
>>     directedGraph.addVertex(2);
>>     System.out.println(directedGraph.addEdge(1,2,s));
>>     System.out.println(directedGraph.addEdge(2,1,s));
>> }
>>
>> The second addEdge call will return false because edge s is already in
>> the graph. What would be a clean way to resolve this issue?
>>
>> Thanks,
>>
>> Joris
>>
>>
>> ------------------------------------------------------------------------------
>> Monitor 25 network devices or servers for free with OpManager!
>> OpManager is web-based network management software that monitors
>> network devices and physical & virtual servers, alerts via email & sms
>> for fault. Monitor 25 devices for free with no restriction. Download now
>> http://ad.doubleclick.net/ddm/clk/292181274;119417398;o
>> _______________________________________________
>> jgrapht-users mailing list
>> jgr...@li...
>> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>>
>>
>

Re: [jgrapht-users] add edge to directed graph in both directions

[Szabolcs B. <bes...@gm...>]

http://jgrapht-users.107614.n3.nabble.com/Mixed-graphs-td4024952.html

Üdvözlettel,

Besenyei Szabolcs

2015-06-23 18:30 GMT+02:00 Joris Kinable <de...@gm...>:

> Say I have a list of directed and undirected edges which I would like to
> represent in the same graph. Every undirected edge (i,j) can be represented
> in a directed graph by adding two arcs: (i,j) and (j,i). The following code
> does however *not* achieve this:
>
> public  static void main(String[] args){
>     DirectedGraph<Integer, String> directedGraph=new
> SimpleDirectedGraph<Integer, String>(String.class);
>     String s="UndirectedEdge";
>     directedGraph.addVertex(1);
>     directedGraph.addVertex(2);
>     System.out.println(directedGraph.addEdge(1,2,s));
>     System.out.println(directedGraph.addEdge(2,1,s));
> }
>
> The second addEdge call will return false because edge s is already in the
> graph. What would be a clean way to resolve this issue?
>
> Thanks,
>
> Joris
>
>
> ------------------------------------------------------------------------------
> Monitor 25 network devices or servers for free with OpManager!
> OpManager is web-based network management software that monitors
> network devices and physical & virtual servers, alerts via email & sms
> for fault. Monitor 25 devices for free with no restriction. Download now
> http://ad.doubleclick.net/ddm/clk/292181274;119417398;o
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

[jgrapht-users] add edge to directed graph in both directions

[Joris K. <de...@gm...>]

Say I have a list of directed and undirected edges which I would like to
represent in the same graph. Every undirected edge (i,j) can be represented
in a directed graph by adding two arcs: (i,j) and (j,i). The following code
does however *not* achieve this:

public  static void main(String[] args){
    DirectedGraph<Integer, String> directedGraph=new
SimpleDirectedGraph<Integer, String>(String.class);
    String s="UndirectedEdge";
    directedGraph.addVertex(1);
    directedGraph.addVertex(2);
    System.out.println(directedGraph.addEdge(1,2,s));
    System.out.println(directedGraph.addEdge(2,1,s));
}

The second addEdge call will return false because edge s is already in the
graph. What would be a clean way to resolve this issue?

Thanks,

Joris

[jgrapht-users] graphbuilder classes?

[Joris K. <de...@gm...>]

Hi,

While browsing the source of jgrapht, I stumbled upon a series of classes
involving a GraphBuilder, dated 12-Jan-2015:
AbstractGraphBuilder.java
DirectedGraphBuilderBase.java
DirectedGraphBuilder.java
DirectedWeightedGraphBuilderBase.java
DirectedWeightedGraphBuilder.java
UndirectedGraphBuilderBase.java
UndirectedGraphBuilder.java
UndirectedWeightedGraphBuilderBase.java
UndirectedWeightedGraphBuilder.java

a. Is there a particular reason I cannot find these classes in the javadoc
linked from jgrapht's website: http://jgrapht.org/javadoc/. Could it be
that the javadoc hasn't been updated with the latest release of jgrapht?

b. Perhaps a bit of a sensitive point: I'm not entirely sure what these
classes add to jgrapht? The majority of methods in these builder classes
were already supported by the default graph classes, or through the Graphs
class. The only new method I can really see is: 'addEdgeChain(V first, V
second, V ... rest). I would recommend to add this method to the Graphs
class, and, unless there's a valid use for these builder classes, consider
to remove them from jgrapht.

Joris

[jgrapht-users] Eclipse

[Manuel S. <spl...@ya...>]

Hi,
I need to customize some algorithms for my own use, I wanted to know how I can import the JGraphT package into Eclipse.
Thank you very much-Manuel

Re: [jgrapht-users] Suggestions for "shortest path with conditional nodes"?

[<org...@io...>]

On 2015-04-13T05:42:02 +0000
"Hoeltzcke, Claus-Dieter (EXTERN: Ethon)"
<ext...@vo...> wrote:

> Hi,
> 
> use masks using the MaskFunctor<> interface.

On 2015-04-13T10:26:30 -0700
John Sichi <js...@gm...> wrote:

> You can use one of the subgraph implementations for this:
> 
> http://jgrapht.org/javadoc/org/jgrapht/graph/Subgraph.html
> http://jgrapht.org/javadoc/org/jgrapht/graph/MaskSubgraph.html
> 

Interesting!

Thanks all, this looks ideal.

M

Re: [jgrapht-users] Suggestions for "shortest path with conditional nodes"?

[John S. <js...@gm...>]

You can use one of the subgraph implementations for this:

http://jgrapht.org/javadoc/org/jgrapht/graph/Subgraph.html
http://jgrapht.org/javadoc/org/jgrapht/graph/MaskSubgraph.html


On Mon, Apr 13, 2015 at 8:00 AM, Rushang Karia <rus...@ho...>
wrote:

> The best way to do this acc to me is to modify the shortest path algorithm.
> Keep a hashset of the nodes which are turned off and simply prune the
> recursion tree.
>
> You can copy paste the source code of the algorithm in a new class.
>
> Let me know if you have questions or if you find a better solution.
>
> -----Original Message-----
> From: org...@io... [mailto:org...@io...]
> Sent: Sunday, April 12, 2015 6:23 AM
> To: jgr...@li...
> Subject: [jgrapht-users] Suggestions for "shortest path with conditional
> nodes"?
>
> Hello.
>
> I'm modelling an abstract network. The network is represented by a simple
> undirected graph (SimpleGraph, specifically). Vertices in the graph
> represent networked devices, and the edges represent hardwired network
> links.
>
> I calculate routes in the network using DijkstraShortestPath. This all
> works
> very well, but I'd now like to model nodes being optionally on/off. If a
> node is "off", then it should be considered a dead end when calculating
> routes. In other words, if a node is "off", then it should be routed around
> if possible.
>
> Does anyone have any suggestions as to how to get the shortest paths when
> some nodes should conditionally be considered "invisible"?
>
> In order to keep the code that deals with graphs simple, and because the
> edges in the graph represent hardwired links, I don't really want to remove
> vertices/edges when a node is switched off - the edges carry a meaning
> beyond simply whether or not a packet can be routed that way.
>
> Right now, I'm leaning towards maintaining two graphs (one with the
> physical
> links, the other with only the active nodes and links), but I'd rather not
> do this as it obviously requires keeping two mutable data structures
> synchronized.
>
> M
>
>
> ----------------------------------------------------------------------------
> --
> BPM Camp - Free Virtual Workshop May 6th at 10am PDT/1PM EDT Develop your
> own process in accordance with the BPMN 2 standard Learn Process modeling
> best practices with Bonita BPM through live exercises
> http://www.bonitasoft.com/be-part-of-it/events/bpm-camp-virtual-
> event?utm_
> source=Sourceforge_BPM_Camp_5_6_15&utm_medium=email&utm_campaign=VA_SF
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>
>
> ------------------------------------------------------------------------------
> BPM Camp - Free Virtual Workshop May 6th at 10am PDT/1PM EDT
> Develop your own process in accordance with the BPMN 2 standard
> Learn Process modeling best practices with Bonita BPM through live
> exercises
> http://www.bonitasoft.com/be-part-of-it/events/bpm-camp-virtual-
> event?utm_
> source=Sourceforge_BPM_Camp_5_6_15&utm_medium=email&utm_campaign=VA_SF
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>

Re: [jgrapht-users] Suggestions for "shortest path with conditional nodes"?

[Rushang K. <rus...@ho...>]

The best way to do this acc to me is to modify the shortest path algorithm.
Keep a hashset of the nodes which are turned off and simply prune the
recursion tree.

You can copy paste the source code of the algorithm in a new class.

Let me know if you have questions or if you find a better solution.

-----Original Message-----
From: org...@io... [mailto:org...@io...] 
Sent: Sunday, April 12, 2015 6:23 AM
To: jgr...@li...
Subject: [jgrapht-users] Suggestions for "shortest path with conditional
nodes"?

Hello.

I'm modelling an abstract network. The network is represented by a simple
undirected graph (SimpleGraph, specifically). Vertices in the graph
represent networked devices, and the edges represent hardwired network
links.

I calculate routes in the network using DijkstraShortestPath. This all works
very well, but I'd now like to model nodes being optionally on/off. If a
node is "off", then it should be considered a dead end when calculating
routes. In other words, if a node is "off", then it should be routed around
if possible.

Does anyone have any suggestions as to how to get the shortest paths when
some nodes should conditionally be considered "invisible"?

In order to keep the code that deals with graphs simple, and because the
edges in the graph represent hardwired links, I don't really want to remove
vertices/edges when a node is switched off - the edges carry a meaning
beyond simply whether or not a packet can be routed that way.

Right now, I'm leaning towards maintaining two graphs (one with the physical
links, the other with only the active nodes and links), but I'd rather not
do this as it obviously requires keeping two mutable data structures
synchronized.

M

----------------------------------------------------------------------------
--
BPM Camp - Free Virtual Workshop May 6th at 10am PDT/1PM EDT Develop your
own process in accordance with the BPMN 2 standard Learn Process modeling
best practices with Bonita BPM through live exercises
http://www.bonitasoft.com/be-part-of-it/events/bpm-camp-virtual- event?utm_
source=Sourceforge_BPM_Camp_5_6_15&utm_medium=email&utm_campaign=VA_SF
_______________________________________________
jgrapht-users mailing list
jgr...@li...
https://lists.sourceforge.net/lists/listinfo/jgrapht-users

[jgrapht-users] Suggestions for "shortest path with conditional nodes"?

[<org...@io...>]

Hello.

I'm modelling an abstract network. The network is represented by a
simple undirected graph (SimpleGraph, specifically). Vertices in the
graph represent networked devices, and the edges represent hardwired
network links.

I calculate routes in the network using DijkstraShortestPath. This
all works very well, but I'd now like to model nodes being optionally
on/off. If a node is "off", then it should be considered a dead end
when calculating routes. In other words, if a node is "off", then it
should be routed around if possible.

Does anyone have any suggestions as to how to get the shortest paths
when some nodes should conditionally be considered "invisible"?

In order to keep the code that deals with graphs simple, and because
the edges in the graph represent hardwired links, I don't really want
to remove vertices/edges when a node is switched off - the edges
carry a meaning beyond simply whether or not a packet can be routed
that way.

Right now, I'm leaning towards maintaining two graphs (one with the
physical links, the other with only the active nodes and links), but
I'd rather not do this as it obviously requires keeping two mutable
data structures synchronized.

M

[jgrapht-users] JGraphT Version 0.9.1 Released

[John S. <js...@gm...>]

This is an incremental release including all contributions since the 0.9.0
release. Starting with this release, JGraphT requires JDK 1.7 to build.

You can see the full change list here:

https://github.com/jgrapht/jgrapht/blob/master/HISTORY.md

Re: [jgrapht-users] 0.9.1 release proposal

[Chris K W. <ch...@we...>]

this is great new, thanks!

> On Apr 5, 2015, at 4:24 PM, John Sichi <js...@gm...> wrote:
> 
> I got some time for it this weekend, so I just published the release in Maven central.  There will be an official announcement once I've completed the non-Maven portion of the release process, but if you get a chance, please test the release via mvn now to make sure everything is correct.
> 
> On Thu, Apr 2, 2015 at 12:25 AM, John Sichi <js...@gm... <mailto:js...@gm...>> wrote:
> I'll try to get it done sooner if possible.
> 
> On Mon, Mar 30, 2015 at 8:03 PM, Chris K Wensel <ch...@we... <mailto:ch...@we...>> wrote:
> I realized a few minutes after hitting send that independently publishing won’t really work. 
> 
> now that JGraphT is in maven central, and If any users are using JGraphT or have made it an explicit dependency, they will end up with two versions in the CLASSPATH, defeating Ivy/Maven’s attempts at resolving versions.
> 
> If end of April is a hard date, we will need to revert to 0.9.0 for our release.
> 
> ckw
> 
> 
>> On Mar 25, 2015, at 3:48 PM, Chris K Wensel <ch...@we... <mailto:ch...@we...>> wrote:
>> 
>> That’s fair.
>> 
>> but to limit the trouble/confusion by our downstream projects (both open-source and commercial Cascading embedders) and to get the most bake time, I’ll need to 'fork and publish' in our public maven repo under a new group name and without the SNAPSHOT classifier.
>> 
>> we did this some years ago before jgrapht was in maven central.
>> 
>> ckw 
>> 
>>> On Mar 25, 2015, at 1:40 PM, John Sichi <js...@gm... <mailto:js...@gm...>> wrote:
>>> 
>>> Sounds good; I'll try to get around to it by end of April.
>>> 
>>> On Mon, Mar 23, 2015 at 8:16 PM, Chris K Wensel <ch...@we... <mailto:ch...@we...>> wrote:
>>> Hi all,
>>> 
>>> The Cascading project would be very interested in a final release of the current 0.9.1 development branch.
>>> 
>>> We are currently relying on a recent patch that has been committed, but in order to make a Cascading 3.0 final release we would need rely on a SNAPSHOT build.
>>> 
>>> So if there aren’t any outstanding blockers on 0.9.1, I would like to move that the current branch be finalized and released.
>>> 
>>> If there are objections or blockers, I’d be happy to help move any issue forward.
>>> 
>>> ckw
>>> 
>>> —
>>> Chris K Wensel
>>> ch...@we... <mailto:ch...@we...>
>>> 
>>> 
>>> 
>>> 
>>> 
>>> ------------------------------------------------------------------------------
>>> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
>>> by Intel and developed in partnership with Slashdot Media, is your hub for all
>>> things parallel software development, from weekly thought leadership blogs to
>>> news, videos, case studies, tutorials and more. Take a look and join the
>>> conversation now. http://goparallel.sourceforge.net/ <http://goparallel.sourceforge.net/>
>>> _______________________________________________
>>> jgrapht-users mailing list
>>> jgr...@li... <mailto:jgr...@li...>
>>> https://lists.sourceforge.net/lists/listinfo/jgrapht-users <https://lists.sourceforge.net/lists/listinfo/jgrapht-users>
>>> 
>>> 
>> 
>> —
>> Chris K Wensel
>> ch...@we... <mailto:ch...@we...>
>> 
>> 
>> 
>> 
>> ------------------------------------------------------------------------------
>> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
>> by Intel and developed in partnership with Slashdot Media, is your hub for all
>> things parallel software development, from weekly thought leadership blogs to
>> news, videos, case studies, tutorials and more. Take a look and join the 
>> conversation now. http://goparallel.sourceforge.net/_______________________________________________ <http://goparallel.sourceforge.net/_______________________________________________>
>> jgrapht-users mailing list
>> jgr...@li... <mailto:jgr...@li...>
>> https://lists.sourceforge.net/lists/listinfo/jgrapht-users <https://lists.sourceforge.net/lists/listinfo/jgrapht-users>
> 
> —
> Chris K Wensel
> ch...@we... <mailto:ch...@we...>
> 
> 
> 
> 
> 
> 

—
Chris K Wensel
ch...@we...

Re: [jgrapht-users] 0.9.1 release proposal

[John S. <js...@gm...>]

I got some time for it this weekend, so I just published the release in
Maven central.  There will be an official announcement once I've completed
the non-Maven portion of the release process, but if you get a chance,
please test the release via mvn now to make sure everything is correct.

On Thu, Apr 2, 2015 at 12:25 AM, John Sichi <js...@gm...> wrote:

> I'll try to get it done sooner if possible.
>
> On Mon, Mar 30, 2015 at 8:03 PM, Chris K Wensel <ch...@we...> wrote:
>
>> I realized a few minutes after hitting send that independently publishing
>> won’t really work.
>>
>> now that JGraphT is in maven central, and If any users are using JGraphT
>> or have made it an explicit dependency, they will end up with two versions
>> in the CLASSPATH, defeating Ivy/Maven’s attempts at resolving versions.
>>
>> If end of April is a hard date, we will need to revert to 0.9.0 for our
>> release.
>>
>> ckw
>>
>>
>> On Mar 25, 2015, at 3:48 PM, Chris K Wensel <ch...@we...> wrote:
>>
>> That’s fair.
>>
>> but to limit the trouble/confusion by our downstream projects (both
>> open-source and commercial Cascading embedders) and to get the most bake
>> time, I’ll need to 'fork and publish' in our public maven repo under a new
>> group name and without the SNAPSHOT classifier.
>>
>> we did this some years ago before jgrapht was in maven central.
>>
>> ckw
>>
>> On Mar 25, 2015, at 1:40 PM, John Sichi <js...@gm...> wrote:
>>
>> Sounds good; I'll try to get around to it by end of April.
>>
>> On Mon, Mar 23, 2015 at 8:16 PM, Chris K Wensel <ch...@we...> wrote:
>>
>>> Hi all,
>>>
>>> The Cascading project would be very interested in a final release of the
>>> current 0.9.1 development branch.
>>>
>>> We are currently relying on a recent patch that has been committed, but
>>> in order to make a Cascading 3.0 final release we would need rely on a
>>> SNAPSHOT build.
>>>
>>> So if there aren’t any outstanding blockers on 0.9.1, I would like to
>>> move that the current branch be finalized and released.
>>>
>>> If there are objections or blockers, I’d be happy to help move any issue
>>> forward.
>>>
>>> ckw
>>>
>>> —
>>> Chris K Wensel
>>> ch...@we...
>>>
>>>
>>>
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Dive into the World of Parallel Programming The Go Parallel Website,
>>> sponsored
>>> by Intel and developed in partnership with Slashdot Media, is your hub
>>> for all
>>> things parallel software development, from weekly thought leadership
>>> blogs to
>>> news, videos, case studies, tutorials and more. Take a look and join the
>>> conversation now. http://goparallel.sourceforge.net/
>>> _______________________________________________
>>> jgrapht-users mailing list
>>> jgr...@li...
>>> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>>>
>>>
>>
>> —
>> Chris K Wensel
>> ch...@we...
>>
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Dive into the World of Parallel Programming The Go Parallel Website,
>> sponsored
>> by Intel and developed in partnership with Slashdot Media, is your hub
>> for all
>> things parallel software development, from weekly thought leadership
>> blogs to
>> news, videos, case studies, tutorials and more. Take a look and join the
>> conversation now.
>> http://goparallel.sourceforge.net/_______________________________________________
>> jgrapht-users mailing list
>> jgr...@li...
>> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>>
>>
>> —
>> Chris K Wensel
>> ch...@we...
>>
>>
>>
>>
>>
>

Re: [jgrapht-users] 0.9.1 release proposal

[John S. <js...@gm...>]

I'll try to get it done sooner if possible.

On Mon, Mar 30, 2015 at 8:03 PM, Chris K Wensel <ch...@we...> wrote:

> I realized a few minutes after hitting send that independently publishing
> won’t really work.
>
> now that JGraphT is in maven central, and If any users are using JGraphT
> or have made it an explicit dependency, they will end up with two versions
> in the CLASSPATH, defeating Ivy/Maven’s attempts at resolving versions.
>
> If end of April is a hard date, we will need to revert to 0.9.0 for our
> release.
>
> ckw
>
>
> On Mar 25, 2015, at 3:48 PM, Chris K Wensel <ch...@we...> wrote:
>
> That’s fair.
>
> but to limit the trouble/confusion by our downstream projects (both
> open-source and commercial Cascading embedders) and to get the most bake
> time, I’ll need to 'fork and publish' in our public maven repo under a new
> group name and without the SNAPSHOT classifier.
>
> we did this some years ago before jgrapht was in maven central.
>
> ckw
>
> On Mar 25, 2015, at 1:40 PM, John Sichi <js...@gm...> wrote:
>
> Sounds good; I'll try to get around to it by end of April.
>
> On Mon, Mar 23, 2015 at 8:16 PM, Chris K Wensel <ch...@we...> wrote:
>
>> Hi all,
>>
>> The Cascading project would be very interested in a final release of the
>> current 0.9.1 development branch.
>>
>> We are currently relying on a recent patch that has been committed, but
>> in order to make a Cascading 3.0 final release we would need rely on a
>> SNAPSHOT build.
>>
>> So if there aren’t any outstanding blockers on 0.9.1, I would like to
>> move that the current branch be finalized and released.
>>
>> If there are objections or blockers, I’d be happy to help move any issue
>> forward.
>>
>> ckw
>>
>> —
>> Chris K Wensel
>> ch...@we...
>>
>>
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Dive into the World of Parallel Programming The Go Parallel Website,
>> sponsored
>> by Intel and developed in partnership with Slashdot Media, is your hub
>> for all
>> things parallel software development, from weekly thought leadership
>> blogs to
>> news, videos, case studies, tutorials and more. Take a look and join the
>> conversation now. http://goparallel.sourceforge.net/
>> _______________________________________________
>> jgrapht-users mailing list
>> jgr...@li...
>> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>>
>>
>
> —
> Chris K Wensel
> ch...@we...
>
>
>
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now.
> http://goparallel.sourceforge.net/_______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>
> —
> Chris K Wensel
> ch...@we...
>
>
>
>
>

Re: [jgrapht-users] 0.9.1 release proposal

[Chris K W. <ch...@we...>]

I realized a few minutes after hitting send that independently publishing won’t really work. 

now that JGraphT is in maven central, and If any users are using JGraphT or have made it an explicit dependency, they will end up with two versions in the CLASSPATH, defeating Ivy/Maven’s attempts at resolving versions.

If end of April is a hard date, we will need to revert to 0.9.0 for our release.

ckw


> On Mar 25, 2015, at 3:48 PM, Chris K Wensel <ch...@we...> wrote:
> 
> That’s fair.
> 
> but to limit the trouble/confusion by our downstream projects (both open-source and commercial Cascading embedders) and to get the most bake time, I’ll need to 'fork and publish' in our public maven repo under a new group name and without the SNAPSHOT classifier.
> 
> we did this some years ago before jgrapht was in maven central.
> 
> ckw 
> 
>> On Mar 25, 2015, at 1:40 PM, John Sichi <js...@gm... <mailto:js...@gm...>> wrote:
>> 
>> Sounds good; I'll try to get around to it by end of April.
>> 
>> On Mon, Mar 23, 2015 at 8:16 PM, Chris K Wensel <ch...@we... <mailto:ch...@we...>> wrote:
>> Hi all,
>> 
>> The Cascading project would be very interested in a final release of the current 0.9.1 development branch.
>> 
>> We are currently relying on a recent patch that has been committed, but in order to make a Cascading 3.0 final release we would need rely on a SNAPSHOT build.
>> 
>> So if there aren’t any outstanding blockers on 0.9.1, I would like to move that the current branch be finalized and released.
>> 
>> If there are objections or blockers, I’d be happy to help move any issue forward.
>> 
>> ckw
>> 
>> —
>> Chris K Wensel
>> ch...@we... <mailto:ch...@we...>
>> 
>> 
>> 
>> 
>> 
>> ------------------------------------------------------------------------------
>> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
>> by Intel and developed in partnership with Slashdot Media, is your hub for all
>> things parallel software development, from weekly thought leadership blogs to
>> news, videos, case studies, tutorials and more. Take a look and join the
>> conversation now. http://goparallel.sourceforge.net/ <http://goparallel.sourceforge.net/>
>> _______________________________________________
>> jgrapht-users mailing list
>> jgr...@li... <mailto:jgr...@li...>
>> https://lists.sourceforge.net/lists/listinfo/jgrapht-users <https://lists.sourceforge.net/lists/listinfo/jgrapht-users>
>> 
>> 
> 
> —
> Chris K Wensel
> ch...@we... <mailto:ch...@we...>
> 
> 
> 
> 
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for all
> things parallel software development, from weekly thought leadership blogs to
> news, videos, case studies, tutorials and more. Take a look and join the 
> conversation now. http://goparallel.sourceforge.net/_______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users

—
Chris K Wensel
ch...@we...

Re: [jgrapht-users] 0.9.1 release proposal

[Chris K W. <ch...@we...>]

That’s fair.

but to limit the trouble/confusion by our downstream projects (both open-source and commercial Cascading embedders) and to get the most bake time, I’ll need to 'fork and publish' in our public maven repo under a new group name and without the SNAPSHOT classifier.

we did this some years ago before jgrapht was in maven central.

ckw 

> On Mar 25, 2015, at 1:40 PM, John Sichi <js...@gm...> wrote:
> 
> Sounds good; I'll try to get around to it by end of April.
> 
> On Mon, Mar 23, 2015 at 8:16 PM, Chris K Wensel <ch...@we... <mailto:ch...@we...>> wrote:
> Hi all,
> 
> The Cascading project would be very interested in a final release of the current 0.9.1 development branch.
> 
> We are currently relying on a recent patch that has been committed, but in order to make a Cascading 3.0 final release we would need rely on a SNAPSHOT build.
> 
> So if there aren’t any outstanding blockers on 0.9.1, I would like to move that the current branch be finalized and released.
> 
> If there are objections or blockers, I’d be happy to help move any issue forward.
> 
> ckw
> 
> —
> Chris K Wensel
> ch...@we... <mailto:ch...@we...>
> 
> 
> 
> 
> 
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for all
> things parallel software development, from weekly thought leadership blogs to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/ <http://goparallel.sourceforge.net/>
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li... <mailto:jgr...@li...>
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users <https://lists.sourceforge.net/lists/listinfo/jgrapht-users>
> 
> 

—
Chris K Wensel
ch...@we...

Re: [jgrapht-users] 0.9.1 release proposal

[John S. <js...@gm...>]

Sounds good; I'll try to get around to it by end of April.

On Mon, Mar 23, 2015 at 8:16 PM, Chris K Wensel <ch...@we...> wrote:

> Hi all,
>
> The Cascading project would be very interested in a final release of the
> current 0.9.1 development branch.
>
> We are currently relying on a recent patch that has been committed, but in
> order to make a Cascading 3.0 final release we would need rely on a
> SNAPSHOT build.
>
> So if there aren’t any outstanding blockers on 0.9.1, I would like to move
> that the current branch be finalized and released.
>
> If there are objections or blockers, I’d be happy to help move any issue
> forward.
>
> ckw
>
> —
> Chris K Wensel
> ch...@we...
>
>
>
>
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

[jgrapht-users] 0.9.1 release proposal

[Chris K W. <ch...@we...>]

Hi all,

The Cascading project would be very interested in a final release of the current 0.9.1 development branch.

We are currently relying on a recent patch that has been committed, but in order to make a Cascading 3.0 final release we would need rely on a SNAPSHOT build.

So if there aren’t any outstanding blockers on 0.9.1, I would like to move that the current branch be finalized and released.

If there are objections or blockers, I’d be happy to help move any issue forward.

ckw

—
Chris K Wensel
ch...@we...

Re: [jgrapht-users] About Cliques

[Joris K. <de...@gm...>]

What exactly are you trying to achieve? According to the documentation, the
getAllMaximalCliques() returns a collection of cliques in the graph:
Collection<Set<V>> So each item in this collection represents a clique.

BronKerboschCliqueFinder cliqueFinder=new BronKerboschCliqueFinder(mygraph);

for(Set<V> clique : cliqueFinder.getAllMaximalCliques()){
  System.out.println("Found a clique with "+clique.size()+" vertices");
}

br,

Joris

On Wed, Mar 4, 2015 at 6:00 PM, fee...@gm... <fee...@gm...>
wrote:

> HI,
>
> I’m looking for help regarding Cliques in graphs. I’ve implemented
>  BronKerboschCliqueFinder but  what I actually want to  obtain is not the
> clique itself... rather its size (number of vertices)…Does anyone know how
> to do this??
>
> I also have the same question but regarding the connectedSets case.
>
> Thanks
>
> Phil
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

[jgrapht-users] About Cliques

[<fee...@gm...>]

HI, 

I’m looking for help regarding Cliques in graphs. I’ve implemented  BronKerboschCliqueFinder but  what I actually want to  obtain is not the clique itself... rather its size (number of vertices)…Does anyone know how to do this??

I also have the same question but regarding the connectedSets case.

Thanks

Phil

[jgrapht-users] Subgraph extracted from Edge/Vertex property

[Armando G. <arm...@me...>]

Hi All,

I'm integrating JGraphT into a reverse logistics cost optimization process by modeling the supply chain as a graph. I'd like to keep my model as general as possible by adding labels to vertices and edges objects. 

My target is to retrieve a sub-graph where the vertices and edges match a supplied label. 

Has anyone tried such a method?

I guess this would need sub classing the graph implementation I'm using rather than writing an algorithm-like class, right?

Armando Garcia
arm...@me...

Re: [jgrapht-users] Please, i need help concerning the SimpleDirectedWeightedGraph Class

[Joris K. <de...@gm...>]

Without seeing more code it's difficult to check what goes wrong. It's very
unlikely that there is a problem in the AbstractBaseGraph implementation.
Check your implementation of your getSlave method first.

Why don't you use the following:
public int nbrDescendants(Site s){
 int i=0;
 Set<Lien> children = graph.incomingEdgesOf(s);
  for(Lien lien: children){
    i+=graph.inDegreeOf(graph.getEdgeTarget(lien));
   }
  return i;
}

Btw, I don't really see the point of this code. If there are 5 incoming
edges for a given site s, then this method will return 5*5=25? Unless I
misunderstood what you mean with "destination vertex of an edge"?

best,

Joris


On Thu, Feb 26, 2015 at 3:39 PM, mokhtar aitbaomar <
ait...@gm...> wrote:

> Hello everyone;
> I have a project in which i need to implement a graph (directed &
> weighted), and i found your work which helped me a lot, but i have a
> problem and i want you to help me out.
> When i create my SimpleDirectedWeightedGraph object "graph" and insret the
> vertexes (type Site in my case) and edges (Type Lien in my case) and apply
> the "incomingEdgesOf(Vertex v)" method to the "graph", i put the result in
> a Set<Lien> and i find that the size (number of incoming edges) is correct
> but when i apply the method again on the resulting set of edges "so that i
> can calculate the number of descendants of an Edge" an exception in thrown
> which says java.lang.NullPointerException .
>
> -------------------------------------------------------------------------------
> Here is my code:
> //CALCULATE THE NUMBER OF DESCENDANTS OF AN EDGE (WHERE SITE S IS THE
> DESTINATION VERTEX) IN THE GRAPH
>     public int nbrDescendants(Site s)
>     {
>         int i=0;
>         Set<Lien> children = graph.incomingEdgesOf(s);
>         Iterator<Lien> it = children.iterator();
>         while(it.hasNext())
>         {
>             i+= graph.incomingEdgesOf(it.next().getSlave()).size();
> //getSlave() returs in my case the destination vertex of the edge
>         }
>         i+=children.size();
>         return i;
>     }
>
> ​-----------------------------------------------------------------------------------​
> ​Here is the error message:
> java.lang.NullPointerException
>     at
> org.jgrapht.graph.AbstractGraph.assertVertexExist(AbstractGraph.java:156)
>     at
> org.jgrapht.graph.AbstractBaseGraph$DirectedSpecifics.getEdgeContainer(AbstractBaseGraph.java:925)
>     at
> org.jgrapht.graph.AbstractBaseGraph$DirectedSpecifics.incomingEdgesOf(AbstractBaseGraph.java:885)
>     at
> org.jgrapht.graph.AbstractBaseGraph.incomingEdgesOf(AbstractBaseGraph.java:411)
>     at inwi.capacite.calcul.Reseau.nbrDescendants(Reseau.java:62)
>     at inwi.capacite.calcul.Rapport.main(Rapport.java:86)
>
> ------------------------------------------------------------------------------------
> And finally i'm not sure but i think that the error comes from the
> AbstractBaseGraph.java file, where i found this:
>
>    */
>         public Set<EE> getUnmodifiableOutgoingEdges()
>         {
>             if (unmodifiableOutgoing == null) {
>                 unmodifiableOutgoing =
> Collections.unmodifiableSet(outgoing);
>             }
>
>             return unmodifiableOutgoing;
>         }
> ​
> ​i think this methode must return Set<E> and not Set<EE>.​
>
>
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>