jgrapht-users — JGraphT user list.

[jgrapht-users] No such vertex in Graph error

[Pallavi K. <kar...@gm...>]

Hi,

Have formed a DefaultDirectedGraph like this:
public DirectedGraph<OWLEntity, DefaultEdge> OWLGraph = new
DefaultDirectedGraph<OWLEntity, DefaultEdge>(DefaultEdge.class);

Have added vertices and edges,
Vertex list is printed as follows:
System.out.println(OWLGraph.vertexSet());

The list shows the vertex:
http://www.semanticweb.org/drkm/ontologies/2017/4/
untitled-ontology-82#Disease

Am constructing all directed paths from the graph as follows:
AllDirectedPaths allPathsGraph = new AllDirectedPaths(toGraph.OWLGraph);

Here, OWLGraph is the default directed graph,

Next, i call getAllPaths from source vertex http://www.semanticweb.
org/drkm/ontologies/2017/4/untitled-ontology-82#Food to destination vertex:
http://www.semanticweb.org/drkm/ontologies/2017/4/
untitled-ontology-82#Disease

The stack trace shows the error as:
java.lang.IllegalArgumentException: no such vertex in graph:
http://www.semanticweb.org/drkm/ontologies/2017/4/
untitled-ontology-82#Disease
at org.jgrapht.graph.AbstractGraph.assertVertexExist(AbstractGraph.java:132)
at org.jgrapht.graph.AbstractBaseGraph.incomingEdgesOf(
AbstractBaseGraph.java:365)
at org.jgrapht.alg.AllDirectedPaths.edgeMinDistancesBackwards(
AllDirectedPaths.java:158)
at org.jgrapht.alg.AllDirectedPaths.getAllPaths(AllDirectedPaths.java:105)
at org.jgrapht.alg.AllDirectedPaths.getAllPaths(AllDirectedPaths.java:69)
at com.mycompany.semanticdistance.SemanticDistance.SemanticDistance(
SemanticDistance.java:161)
at com.mycompany.semanticdistance.SemanticDistance.loadOntology(
SemanticDistance.java:136)
at com.mycompany.semanticdistance.SemanticDistance.main(
SemanticDistance.java:84)

May i know how to resolve this?
Kindly help.
Thanks much in advance

-- 
Regs, Pallavi

Linkedin Profile : http://in.linkedin.com/pub/pallavi-karanth/b/b71/26b

“Your time is limited, don’t waste it living someone else’s life. Don’t be
trapped by dogma, which is living the result of other people’s thinking.
Don’t let the noise of other opinions drown your own inner voice. And most
important, have the courage to follow your heart and intuition, they
somehow already know what you truly want to become. Everything else is
secondary.”

-- Steve Jobs

Re: [jgrapht-users] Why are getSource() and getTarget() protected?

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

Applications should use getEdgeSource and getEdgeTarget from the graph
interface rather than accessing these directly.  So they are protected in
order to prevent accidental use.

On Jan 6, 2018 5:04 PM, "Victor Porton" <po...@na...> wrote:

> Why in
> http://jgrapht.org/javadoc/org/jgrapht/graph/DefaultEdge.html
> getSource() and getTarget() are protected (rather than public)?
>
> I don't see any sense of them being non-public.
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

[jgrapht-users] Why are getSource() and getTarget() protected?

[Victor P. <po...@na...>]

Why in
http://jgrapht.org/javadoc/org/jgrapht/graph/DefaultEdge.html
getSource() and getTarget() are protected (rather than public)?

I don't see any sense of them being non-public.

Re: [jgrapht-users] DefaultEdge vs Void

[J K. <j.k...@gm...>]

Yes that would work. In Java, any class comes with a constructor, whether
you explicitly define it or not. Of course, you would still have to specify
a EdgeFactory which creates new instances of your edges. Aside from that,
it wouldn't really provide you with a benefit. The edge source and target
are stored on the edge. The default types, DefaultEdge or
DefaultWeightedEdge extend IntrustiveEdge. Even if your provide your own
custom edge type which does not extend IntrusiveEdge, under the hood a
Map<YourEdgeType, IntrusiveEdge> will be used which maps your special
YourEdgeType edge  to a newly created IntrusiveEdge. So in summary, you
won't gain anything.

For reference, have a look at AbstractBaseGraph, IntrusiveEdgesSpecifics
(and one of its implementations UniformIntrusiveEdgesSpecifics).

On Sat, Jan 6, 2018 at 1:51 PM, Victor Porton <po...@na...> wrote:

> But what if I write the following?
>
> class MyVoid {
> // empty
> }
>
> Graph<T,MyVoid>
>
> Will this instantiation "work"?
>
> I ask because it seems to me to store source and target in each edge is
> redundant, because in context of its usage the source and target is known.
> I ask therefore, if we can use an empty class like MyVoid to avoid storing
> source and destination.
>
> On Sat, 2018-01-06 at 19:56 +0200, Dimitrios Michail wrote:
>
> Not really. Void is just a placeholder type which cannot be instantiated.
> The type you provide for the edges
> must be a class that can be instantiated.
>
> Read about Java Generics in order to understand its purpose (
> https://docs.oracle.com/javase/tutorial/java/generics/index.html).
>
> There are plenty of examples of graph instantiations in the JGraphT test
> cases.
>
> --Dimitrios
>
>
> On Sat, Jan 6, 2018 at 4:14 PM, Victor Porton <po...@na...> wrote:
>
> What is the difference between Graph<T,DefaultEdge> and Graph<T,Void>?
>
> Does the second work?
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>
> _______________________________________________
> jgrapht-users mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

Re: [jgrapht-users] DefaultEdge vs Void

[Victor P. <po...@na...>]

But what if I write the following?

class MyVoid {
    // empty
}

Graph<T,MyVoid>

Will this instantiation "work"?

I ask because it seems to me to store source and target in each edge is
redundant, because in context of its usage the source and target is
known. I ask therefore, if we can use an empty class like MyVoid to
avoid storing source and destination.

On Sat, 2018-01-06 at 19:56 +0200, Dimitrios Michail wrote:
> Not really. Void is just a placeholder type which cannot be
> instantiated. The type you provide for the edgesmust be a class that
> can be instantiated.
> 
> Read about Java Generics in order to understand its purpose (https://
> docs.oracle.com/javase/tutorial/java/generics/index.html).
> 
> There are plenty of examples of graph instantiations in the JGraphT
> test cases.
> 
> --Dimitrios
> 
> 
> 
> On Sat, Jan 6, 2018 at 4:14 PM, Victor Porton <po...@na...>
> wrote:
> > What is the difference between Graph<T,DefaultEdge> and
> > Graph<T,Void>?
> > 
> > 
> > 
> > Does the second work?
> > 
> > 
> > 
> > -----------------------------------------------------------------
> > -------------
> > 
> > Check out the vibrant tech community on one of the world's most
> > 
> > engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> > 
> > _______________________________________________
> > 
> > jgrapht-users mailing list
> > 
> > jgr...@li...
> > 
> > https://lists.sourceforge.net/lists/listinfo/jgrapht-users
> > 
> 
> -------------------------------------------------------------------
> -----------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot__________
> _____________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users

Re: [jgrapht-users] DefaultEdge vs Void

[Dimitrios M. <dim...@gm...>]

Not really. Void is just a placeholder type which cannot be instantiated.
The type you provide for the edges
must be a class that can be instantiated.

Read about Java Generics in order to understand its purpose (
https://docs.oracle.com/javase/tutorial/java/generics/index.html).

There are plenty of examples of graph instantiations in the JGraphT test
cases.

--Dimitrios


On Sat, Jan 6, 2018 at 4:14 PM, Victor Porton <po...@na...> wrote:

> What is the difference between Graph<T,DefaultEdge> and Graph<T,Void>?
>
> Does the second work?
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>

[jgrapht-users] DefaultEdge vs Void

[Victor P. <po...@na...>]

What is the difference between Graph<T,DefaultEdge> and Graph<T,Void>?

Does the second work?

Re: [jgrapht-users] Implement composition of graphs

[J K. <j.k...@gm...>]

So if edge (x,y) exists in S and edge (y,z) exists in T, then (x,z) must
exist in U=S*T.
The following simple algorithm would work:

for edge (x,y) in S.edgeSet():
    for z in T.neighborsOf(y):
        if (x != z):
            U.addEdge(x,z)

Consider using the following function:
Graphs.addEdgeWithVertices(x,z)


On Thu, Dec 28, 2017 at 12:38 PM, Victor Porton <po...@na...> wrote:

> Let R and S be directed graphs.
>
> I will denote xTy iff there is an edge from x to y (for digraph T).
>
> Now I want to calculate the graph R*S defined by the formula:
>
> x(R*S)z <=> exists y: (xSy and yRz)
>
> It seems that this is not implemented in JGraphT.
>
> It seems that I will implement this in near future (unless somebody does
> it quicker than me [image: :-)]).
>
> On Thu, 2017-12-28 at 12:19 -0500, J Kinable wrote:
>
> You need to be a bit more explicit as what you want/need. I have no idea
> what you mean by "composition of binary relations as digraph". There is no
> representation of "binary relations". Obviously, there are plenty of
> different digraphs and algorithms available. Many algorithms can be easily
> implemented using the building blocks that are already present in the
> library.
>
> On Thu, Dec 28, 2017 at 8:55 AM, Victor Porton <po...@na...> wrote:
>
> I need composition of digraphs (I mean composition of binary relations
> represented as digraphs.) in my project. It seems that JGraphT is
> missing this operation.
>
> Is it really missing?
>
> If it is missing I want to implement it.
>
> I remind that composition is defined by the formula
>
> x(R*S)z <=> exists y: (xSy and yRz)
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

Re: [jgrapht-users] Implement composition of graphs

[Victor P. <po...@na...>]

Let R and S be directed graphs.

I will denote xTy iff there is an edge from x to y (for digraph T).

Now I want to calculate the graph R*S defined by the formula:

x(R*S)z <=> exists y: (xSy and yRz)

It seems that this is not implemented in JGraphT.

It seems that I will implement this in near future (unless somebody
does it quicker than me ).

On Thu, 2017-12-28 at 12:19 -0500, J Kinable wrote:
> You need to be a bit more explicit as what you want/need. I have no
> idea what you mean by "composition of binary relations as digraph".
> There is no representation of "binary relations". Obviously, there
> are plenty of different digraphs and algorithms available. Many
> algorithms can be easily implemented using the building blocks that
> are already present in the library. 
> On Thu, Dec 28, 2017 at 8:55 AM, Victor Porton <po...@na...>
> wrote:
> > I need composition of digraphs (I mean composition of binary
> > relations
> > 
> > represented as digraphs.) in my project. It seems that JGraphT is
> > 
> > missing this operation.
> > 
> > 
> > 
> > Is it really missing?
> > 
> > 
> > 
> > If it is missing I want to implement it.
> > 
> > 
> > 
> > I remind that composition is defined by the formula
> > 
> > 
> > 
> > x(R*S)z <=> exists y: (xSy and yRz)
> > 
> > 
> > 
> > -----------------------------------------------------------------
> > -------------
> > 
> > Check out the vibrant tech community on one of the world's most
> > 
> > engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> > 
> > _______________________________________________
> > 
> > jgrapht-users mailing list
> > 
> > jgr...@li...
> > 
> > https://lists.sourceforge.net/lists/listinfo/jgrapht-users
> > 

Re: [jgrapht-users] Implement composition of graphs

[J K. <j.k...@gm...>]

You need to be a bit more explicit as what you want/need. I have no idea
what you mean by "composition of binary relations as digraph". There is no
representation of "binary relations". Obviously, there are plenty of
different digraphs and algorithms available. Many algorithms can be easily
implemented using the building blocks that are already present in the
library.

On Thu, Dec 28, 2017 at 8:55 AM, Victor Porton <po...@na...> wrote:

> I need composition of digraphs (I mean composition of binary relations
> represented as digraphs.) in my project. It seems that JGraphT is
> missing this operation.
>
> Is it really missing?
>
> If it is missing I want to implement it.
>
> I remind that composition is defined by the formula
>
> x(R*S)z <=> exists y: (xSy and yRz)
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>

Re: [jgrapht-users] Composition and inverse of a graph

[J K. <j.k...@gm...>]

I have no idea about whether you should use Jung or jgrapht, but as far as
I know, development of Jung stopped several years ago.

1. Check existence of a path between two vertices.
Use of of the (many) shortest path algorithms, or a connectivity inspector.

2. Invert (replace every edge with the opposite edge) a directed graph.
Use http://jgrapht.org/javadoc/org/jgrapht/graph/EdgeReversedGraph.html

3. No idea what you want.

br,

Joris Kinable


On Wed, Dec 27, 2017 at 1:27 PM, Victor Porton <po...@na...> wrote:

> I am implementing this ambitious free software in Java:
>
> https://en.wikiversity.org/wiki/Automatic_transformation_of_XML_namespaces
>
> I need a graph library. I am now choosing JUNG vs JGraphT.
>
> Now I have the following requirements:
>
> 1. Check existence of a path between two vertices.
>
> 2. Invert (replace every edge with the opposite edge) a directed graph.
>
> 3. Calculate "mathematical" relational composition of two binary relations
> represented as directed graphs.
>
> My requirements may expand in the future.
>
> Please explain how to do each of my requirements in both JUNG and JGraphT
> and help to choose between these two.
>
> Thanks in advance.
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

[jgrapht-users] Implement composition of graphs

[Victor P. <po...@na...>]

I need composition of digraphs (I mean composition of binary relations
represented as digraphs.) in my project. It seems that JGraphT is
missing this operation.

Is it really missing?

If it is missing I want to implement it.

I remind that composition is defined by the formula

x(R*S)z <=> exists y: (xSy and yRz)

[jgrapht-users] Composition and inverse of a graph

[Victor P. <po...@na...>]

I am implementing this ambitious free software in Java:


https://en.wikiversity.org/wiki/Automatic_transformation_of_XML_namespa
ces


I need a graph library. I am now choosing JUNG vs JGraphT.


Now I have the following requirements:


1. Check existence of a path between two vertices.


2. Invert (replace every edge with the opposite edge) a directed graph.


3. Calculate "mathematical" relational composition of two binary
relations represented as directed graphs.


My requirements may expand in the future.


Please explain how to do each of my requirements in both JUNG and
JGraphT and help to choose between these two.


Thanks in advance.

Re: [jgrapht-users] Shortest path from a source

[Adib R. <adi...@gm...>]

Thanks. Actually, I realized that I should use labeled graph for my case
which jGrapht supports properly.



Thanks,
Adib

On Mon, Nov 27, 2017 at 10:52 AM, J Kinable <j.k...@gm...> wrote:

> Without additional information, your question is impossible to answer. We
> cannot see how your graph is constructed, or what the definition of your
> TopoVertex is.
> Common issues:
> 1. Instead of an undirected graph, you create a *directed* graph with an
> edge from a to b. Then you query the shortest path from b to a, but there
> is no such path.
> 2. You did not implement the hashcode/equals functions, see:
> https://github.com/jgrapht/jgrapht/wiki/EqualsAndHashCode
>
> br,
>
> Joris
>
> On Sun, Nov 26, 2017 at 10:12 PM, Adib Rastegarnia <
> adi...@gm...> wrote:
>
>> Hello,
>>
>> I created a graph using jGrapht library which is based on a network
>> topology and printed the results as follows:
>>
>> of:0000000000000191:3,of:0000000000000003:1
>> of:0000000000000191:2,of:0000000000000002:1
>> of:0000000000000192:3,of:0000000000000003:2
>> of:0000000000000192:2,of:0000000000000002:2
>> of:0000000000000191:1,of:0000000000000001:1
>> of:0000000000000192:1,of:0000000000000001:2
>> of:0000000000000003:1,of:0000000000000191:3
>> of:0000000000000003:2,of:0000000000000192:3
>> of:0000000000000002:1,of:0000000000000191:2
>> of:0000000000000001:1,of:0000000000000191:1
>> of:0000000000000002:2,of:0000000000000192:2
>> of:0000000000000001:2,of:0000000000000192:1
>>
>> Each vertex in my graph is defined as a device ID and a port number. I
>> tried to run a shortest path algorithm to find the shortest path between
>> two end points as follows:
>>
>> DijkstraShortestPath<TopoVertex, TopoEdge> dijkstraAlg =
>>         new DijkstraShortestPath<>(topoStore.getDirectedGraph());
>>
>> TopoVertex src = new TopoVertex("of:0000000000000001", "2");
>> TopoVertex dst = new TopoVertex("of:0000000000000002", "2");
>> ShortestPathAlgorithm.SingleSourcePaths<TopoVertex, TopoEdge> iPaths = dijkstraAlg.getPaths(src);
>> System.out.println(iPaths.getPath(dst) + "\n");
>>
>> It prints null but there is a path between these two end points.
>>
>> Any suggestions?
>>
>> Thanks,
>>
>> Adib
>>
>>
>>
>>
>>
>> ------------------------------------------------------------
>> ------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>> _______________________________________________
>> jgrapht-users mailing list
>> jgr...@li...
>> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>>
>>
>

Re: [jgrapht-users] Shortest path from a source

[J K. <j.k...@gm...>]

Without additional information, your question is impossible to answer. We
cannot see how your graph is constructed, or what the definition of your
TopoVertex is.
Common issues:
1. Instead of an undirected graph, you create a *directed* graph with an
edge from a to b. Then you query the shortest path from b to a, but there
is no such path.
2. You did not implement the hashcode/equals functions, see:
https://github.com/jgrapht/jgrapht/wiki/EqualsAndHashCode

br,

Joris

On Sun, Nov 26, 2017 at 10:12 PM, Adib Rastegarnia <
adi...@gm...> wrote:

> Hello,
>
> I created a graph using jGrapht library which is based on a network
> topology and printed the results as follows:
>
> of:0000000000000191:3,of:0000000000000003:1
> of:0000000000000191:2,of:0000000000000002:1
> of:0000000000000192:3,of:0000000000000003:2
> of:0000000000000192:2,of:0000000000000002:2
> of:0000000000000191:1,of:0000000000000001:1
> of:0000000000000192:1,of:0000000000000001:2
> of:0000000000000003:1,of:0000000000000191:3
> of:0000000000000003:2,of:0000000000000192:3
> of:0000000000000002:1,of:0000000000000191:2
> of:0000000000000001:1,of:0000000000000191:1
> of:0000000000000002:2,of:0000000000000192:2
> of:0000000000000001:2,of:0000000000000192:1
>
> Each vertex in my graph is defined as a device ID and a port number. I
> tried to run a shortest path algorithm to find the shortest path between
> two end points as follows:
>
> DijkstraShortestPath<TopoVertex, TopoEdge> dijkstraAlg =
>         new DijkstraShortestPath<>(topoStore.getDirectedGraph());
>
> TopoVertex src = new TopoVertex("of:0000000000000001", "2");
> TopoVertex dst = new TopoVertex("of:0000000000000002", "2");
> ShortestPathAlgorithm.SingleSourcePaths<TopoVertex, TopoEdge> iPaths = dijkstraAlg.getPaths(src);
> System.out.println(iPaths.getPath(dst) + "\n");
>
> It prints null but there is a path between these two end points.
>
> Any suggestions?
>
> Thanks,
>
> Adib
>
>
>
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

[jgrapht-users] Shortest path from a source

[Adib R. <adi...@gm...>]

Hello,

I created a graph using jGrapht library which is based on a network
topology and printed the results as follows:

of:0000000000000191:3,of:0000000000000003:1
of:0000000000000191:2,of:0000000000000002:1
of:0000000000000192:3,of:0000000000000003:2
of:0000000000000192:2,of:0000000000000002:2
of:0000000000000191:1,of:0000000000000001:1
of:0000000000000192:1,of:0000000000000001:2
of:0000000000000003:1,of:0000000000000191:3
of:0000000000000003:2,of:0000000000000192:3
of:0000000000000002:1,of:0000000000000191:2
of:0000000000000001:1,of:0000000000000191:1
of:0000000000000002:2,of:0000000000000192:2
of:0000000000000001:2,of:0000000000000192:1

Each vertex in my graph is defined as a device ID and a port number. I
tried to run a shortest path algorithm to find the shortest path between
two end points as follows:

DijkstraShortestPath<TopoVertex, TopoEdge> dijkstraAlg =
        new DijkstraShortestPath<>(topoStore.getDirectedGraph());

TopoVertex src = new TopoVertex("of:0000000000000001", "2");
TopoVertex dst = new TopoVertex("of:0000000000000002", "2");
ShortestPathAlgorithm.SingleSourcePaths<TopoVertex, TopoEdge> iPaths =
dijkstraAlg.getPaths(src);
System.out.println(iPaths.getPath(dst) + "\n");

It prints null but there is a path between these two end points.

Any suggestions?

Thanks,

Adib

[jgrapht-users] Fwd: [jgrapht-announce] JGraphT version 1.1 released

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

We've just merged the complete removal of the deprecated
undirected/directed/weighted graph interfaces into trunk.  So, apps which
depend on them (that is, most apps which use JGraphT) will need to be
updated in order to be able to use versions of JGraphT past the latest
(1.1) release.

Joris, thanks as always for the release work!

---------- Forwarded message ----------
From: J Kinable <j.k...@gm...>
Date: Tue, Nov 14, 2017 at 4:58 AM
Subject: [jgrapht-announce] JGraphT version 1.1 released
To: jgr...@li...,
jgr...@li...


Hi all,

JGraphtT version 1.1 has been released! As always, thank you for your
contributions. This release significantly reduces the number of different
Graph interfaces: the interfaces UndirectedGraph, DirectedGraph and
WeightedGraph have been deprecated. Instead, their functionality is
captured in one, simple, clean *Graph* interface. This makes it much easier
to implement algorithms, methods and classes which can work with various
different graph types.
For a full list of changes, refer to https://github.com/jgrapht/jgrapht
/blob/master/HISTORY.md

cheers,

Joris Kinable

------------------------------------------------------------
------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
jgrapht-announce mailing list
jgr...@li...
https://lists.sourceforge.net/lists/listinfo/jgrapht-announce

[jgrapht-users] JGraphT version 1.1 released

[J K. <j.k...@gm...>]

Hi all,

JGraphtT version 1.1 has been released! As always, thank you for your
contributions. This release significantly reduces the number of different
Graph interfaces: the interfaces UndirectedGraph, DirectedGraph and
WeightedGraph have been deprecated. Instead, their functionality is
captured in one, simple, clean *Graph* interface. This makes it much easier
to implement algorithms, methods and classes which can work with various
different graph types.
For a full list of changes, refer to https://github.com/jgrapht/jgrapht
/blob/master/HISTORY.md

cheers,

Joris Kinable

Re: [jgrapht-users] Extracting parts of the graph

[J K. <j.k...@gm...>]

Currently there's no build-in function which can do that automatically for
you. We are currently making this functionality available somewhere in the
next few weeks, see Pull Request:
https://github.com/jgrapht/jgrapht/pull/423

In the mean time, to get the desired behavior, you could implement your own
DepthFirstSearch iterator. Or more conveniently, just copy mine:
https://github.com/jkinable/jgrapht/blob/95daef5a6826818d00e113ac610415a763c5aee0/jgrapht-core/src/main/java/org/jgrapht/traverse/DepthFirstIterator.java

This iterator has a new function getDepth(V v) which allows you to query
the depth of a specific node (after the DFS completes, or after the node
has been visited). So in order to query all nodes at a certain depth, you
could write an additional function:
//cache the nodes at a given depth
Map<Integer,Set<V>> depthMap=new HashMap<>();
for(V v : graph.vertexSet()){
 int depth=dfs.getDepth(v);
 if(!depthMap.containsKey(depth))
   depthMap.put(depth, new HashSet<V>());
 depthMap.get(depth).add(v);
}
Now you can easily query all nodes at a given depth: depthMap.get(3);

On Thu, Sep 14, 2017 at 1:00 PM, Bobi Adji Andov <bob...@gm...>
wrote:

> Hello all,
>
> I am a new user to JGraphT and as every other newbie I am running into
> ideas that I would like to solve with this library, but I am still not sure
> how to do that.
>
> I am building a directed graph and I am wandering is there a way to get
> all the nodes at certain depth (starting from a vertex). For example
> something like get all nodes at depth 3 from certain vertex.
>
> Thanks in advance,
> Slobodan
>
> --
>
>
>
> *Slobodan Adji-Andovskype:bobiadziandov*
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

[jgrapht-users] Extracting parts of the graph

[Bobi A. A. <bob...@gm...>]

Hello all,

I am a new user to JGraphT and as every other newbie I am running into
ideas that I would like to solve with this library, but I am still not sure
how to do that.

I am building a directed graph and I am wandering is there a way to get all
the nodes at certain depth (starting from a vertex). For example something
like get all nodes at depth 3 from certain vertex.

Thanks in advance,
Slobodan

-- 



*Slobodan Adji-Andovskype:bobiadziandov*

[jgrapht-users] jgrapht release 1.1 pre-announcement

[J K. <j.k...@gm...>]

We are in the process of finalizing the next release of jgrapht, version
1.1. The release is scheduled for sometime this month.
This release contains some major changes, including various simplifications
to the graph interfaces. Moreover, several algorithms have been
re-implemented, thereby adapting them to accept both directed and
undirected graphs, while at the same time improving their
overall performance.

If you have any last-minute changes you would like to make available for
the upcoming release, then this is the time to submit them (pull request or
create a new issue). Similarly, if you still have open PRs lingering
around, please fix any comments by the reviewers.


br,

Joris Kinable

Re: [jgrapht-users] Isomorphic Graphs - VF2

[J K. <j.k...@gm...>]

Please address your replies to the entire list, not to me in person: this
increases the likelihood that the original author of the algorithm chimes
in.

I'm unfortunately not familiar with the VF2 algorithm, or its
implementation in jgrapht. I'd recommend to repeat your experiment. Keep
your experiment as simple as possible to eliminate potential (common)
causes of bugs. E.g.:
1. use Graph<Integer, DefaultEdge> g1=new SimpleGraph<>(DefaultEdge.class);
for both graphs (instead of directed graphs which are cast to undirected
graphs).
2. use Integer as vertex data type, not String.

Finally:
1. Since checking whether 2 graphs are isomorphic is generally hard, please
verify whether VF2 is a complete (exact) algorithm.
2. Provide a mapping from the vertices of g1 to the vertices of g2 to prove
that the graphs are indeed isomorphic.

If the graphs are indeed isomorphic, the algorithm is exact, and the
algorithm returns false, please feel free to file a bug report using our
issue tracker:
https://github.com/jgrapht/jgrapht/issues
Alternatively, if you are convinced that this is a bug, feel free to have a
look at the source code (https://github.com/jgrapht/jgrapht) . The
algorithm seems pretty short, so it may not be very hard to track down
what's going wrong.

thanks,

Joris Kinable


On Thu, Aug 17, 2017 at 2:37 PM, murad tukan <mur...@gm...> wrote:

> Dear Joris,
>
> I have attached bellow the code for checking the isomorphism:
>
>
> public class testing3 {
>
>     private static DefaultDirectedGraph<String, DefaultEdge> graph1,
> graph2;
>     private static IsomorphismInspector<String, DefaultEdge> inspector;
>
>
>     public static void main(String[] args)
>     {
>
>
>         graph1 = new DefaultDirectedGraph<String,
> DefaultEdge>(DefaultEdge.class);
>         graph2 = new DefaultDirectedGraph<String,
> DefaultEdge>(DefaultEdge.class);
>
>         // graph1
>         for(int i = 1; i <= 6; ++i)
>         {
>             graph1.addVertex(Integer.toString(i));
>             graph2.addVertex(Integer.toString(i));
>         }
>
>         graph1.addEdge("1", "2");
>         graph1.addEdge("2", "1");
>         graph1.addEdge("1", "3");
>         graph1.addEdge("3", "1");
>         graph1.addEdge("1", "6");
>         graph1.addEdge("6", "1");
>         graph1.addEdge("2", "4");
>         graph1.addEdge("4", "2");
>         graph1.addEdge("2", "5");
>         graph1.addEdge("5", "2");
>         graph1.addEdge("3", "4");
>         graph1.addEdge("4", "3");
>         graph1.addEdge("3", "5");
>         graph1.addEdge("5", "3");
>         graph1.addEdge("4", "6");
>         graph1.addEdge("6", "4");
>         graph1.addEdge("5", "6");
>         graph1.addEdge("6", "5");
>
>
>         // graph2
>         graph2.addEdge("1", "2");
>         graph2.addEdge("2", "1");
>         graph2.addEdge("1", "3");
>         graph2.addEdge("3", "1");
>         graph2.addEdge("1", "6");
>         graph2.addEdge("6", "1");
>         graph2.addEdge("2", "3");
>         graph2.addEdge("3", "2");
>         graph2.addEdge("2", "4");
>         graph2.addEdge("4", "2");
>         graph2.addEdge("3", "5");
>         graph2.addEdge("5", "3");
>         graph2.addEdge("4", "5");
>         graph2.addEdge("5", "4");
>         graph2.addEdge("4", "6");
>         graph2.addEdge("6", "4");
>         graph2.addEdge("5", "6");
>         graph2.addEdge("6", "5");
>
>         UndirectedGraph<String, DefaultEdge> g1 = new
> AsUndirectedGraph<String, DefaultEdge>(graph1);
>         UndirectedGraph<String, DefaultEdge> g2 = new
> AsUndirectedGraph<String, DefaultEdge>(graph2);
>
>         inspector = new VF2GraphIsomorphismInspector<String,
> DefaultEdge>(g1,g2);
>         System.err.println(inspector.isomorphismExists());
>     }
>
> }
>
> For such graph, the result should be true rather than false (Also i
> checked using MATLAB graph isomorphism solver and it returned true).
>
> To get more intuition why the two graphs are isomorphic, see the actual
> graphs:
>
> [image: Inline image 2]
>
>
> Please advise.
>
> Thanks in advance,
> M.T.
>
> P.s. Also tried using SimpleGraph and still got false in term of
> isomorphic or not.
>
> On Thu, Aug 17, 2017 at 7:48 AM, J Kinable <j.k...@gm...> wrote:
>
>> Hi,
>>
>> Can you be a bit more specific? What is the graph you are testing on? Can
>> you show a code snippet. Preferably provide a small working example, and
>> describe the expected output.
>>
>> br,
>>
>> Joris Kinable
>>
>> On Wed, Aug 16, 2017 at 9:59 PM, murad tukan <mur...@gm...> wrote:
>>
>>> Dear sir/madam,
>>>
>>> I have tried your vf2 algorithm a.k.a. VF2GraphIsomorphismInspector on
>>> the following graphs by replacing each edge by two directional edges such
>>> that each is in opposite direction in order to obtain a directed graph.
>>> The algorithm returned false when it should have returned true.
>>>
>>> Is there a bug in the implementation of VF2 or I misrepresented the
>>> graphs?
>>>
>>> Please advise as soon as possible.
>>>
>>> Thanks in advance,
>>> M.T.
>>>
>>> ------------------------------------------------------------
>>> ------------------
>>> Check out the vibrant tech community on one of the world's most
>>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>>> _______________________________________________
>>> jgrapht-users mailing list
>>> jgr...@li...
>>> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>>>
>>>
>>
>

Re: [jgrapht-users] Isomorphic Graphs - VF2

[J K. <j.k...@gm...>]

Hi,

Can you be a bit more specific? What is the graph you are testing on? Can
you show a code snippet. Preferably provide a small working example, and
describe the expected output.

br,

Joris Kinable

On Wed, Aug 16, 2017 at 9:59 PM, murad tukan <mur...@gm...> wrote:

> Dear sir/madam,
>
> I have tried your vf2 algorithm a.k.a. VF2GraphIsomorphismInspector on the
> following graphs by replacing each edge by two directional edges such that
> each is in opposite direction in order to obtain a directed graph.
> The algorithm returned false when it should have returned true.
>
> Is there a bug in the implementation of VF2 or I misrepresented the graphs?
>
> Please advise as soon as possible.
>
> Thanks in advance,
> M.T.
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> jgrapht-users mailing list
> jgr...@li...
> https://lists.sourceforge.net/lists/listinfo/jgrapht-users
>
>

[jgrapht-users] Isomorphic Graphs - VF2

[murad t. <mur...@gm...>]

Dear sir/madam,

I have tried your vf2 algorithm a.k.a. VF2GraphIsomorphismInspector on the
following graphs by replacing each edge by two directional edges such that
each is in opposite direction in order to obtain a directed graph.
The algorithm returned false when it should have returned true.

Is there a bug in the implementation of VF2 or I misrepresented the graphs?

Please advise as soon as possible.

Thanks in advance,
M.T.

[jgrapht-users] refreshing the view

[Michał U. <mic...@op...>]

ello I've created graph using JGraphT library g = new ListenableUndirectedWeightedGraph <String, MyEdge>(MyEdge.class); graphAdapter = new JGraphXAdapter<String, MyEdge>(g); layout = new mxOrganicLayout(graphAdapter); layout.execute(graphAdapter.getDefaultParent()); mxGraphComponent component = new mxGraphComponent(graphAdapter); component.setPreferredSize(new Dimension(dim.width - 50, dim.height - 200)); add(component);
and I want to add dynamically a new vertex after pushing the button
@Override public void actionPerformed(ActionEvent e) { String a="1"; String b="2"; g.addVertex(a); g.addVertex(b); g.addEdge(a,b); } public static class MyEdge extends DefaultWeightedEdge { //weight @Override public String toString() { return String.valueOf(getWeight()); } }
how can I refresh the view?